Complete native numbered-save round trips

docs/engine-re.md

@@ -1754,8 +1754,9 @@ live surfaces and retained objects, and applies the 20-byte surface records' exp
 load first replaces each serialized mutable bank prefix while preserving initialization-authored cells
 beyond its count, restores history/gfx and retained audio, unwinds the obsolete managed call chain, runs
 `CALLBACK_LOAD.BIN` when
-the mounted script provider resolves it, starts the saved root at its `0xae` rendezvous, recursively
+the mounted script provider resolves it, loads each saved script at its ordinary entry so its frame-local
-reconstructs child frames, resumes parents after their saved T2 call sites,
+prologue runs, lets that script reach its own `0xae` rendezvous, recursively reconstructs child frames,
+resumes parents after their saved T2 call sites,
 and finally resumes the terminal frame at its T1 boundary. Successful `0x19e` also flushes shared
 `SAVE.DAT`/`RT.DAT`, matching `context_state_serialize`.
 
@@ -1791,6 +1792,11 @@ separately gated by both `set:CreateObject` and `set:AutoFreeTex`; the registere
 path. Opcode `0x259` is also a real lifecycle operation:
 `op_0x259_script_entry_clear_surface_persistence_flags@0x417660` clears record `+0x08` and `+0x0c`
 across both native tables at every script entry.
+Opcode `0x258` then declares the exceptions:
+`op_0x258_set_surface_persistence_flags@0x4250a0` calls
+`gfx_surface_set_persistence_flags@0x4159c0`, which replaces record `+0x08` from flags bit 0 and
+record `+0x0c` from bit 1 in both tables. The port models bit 0 because layout-3 restoration consumes
+that reload policy; the adjacent bit-1 consumer remains unnamed.
 Installed slot 15 records resource `0x3383` with reload flag zero. SYSTEM4 loads that atlas before the
 save UI, and BUNKI cuts its reusable choice-box corners, borders, and winged top ornament from it. The
 port formerly released all 1,000 host surfaces unconditionally, so post-load BUNKI retained its black
@@ -1800,6 +1806,67 @@ native all-release branch. Port-authored saves emit the exact resource/reload/cr
 `-1` for unused records, and opcode `0x259` clears the modeled reload policy without releasing textures.
 The corrected native functions are renamed/commented in the saved `/v2` image.
 
+#### Port-authored round trips: restored frame boundary and retained-object encoding (2026-07-24)
+
+An immediate `installed SAVE00 -> port SAVE03` round trip isolated two independent writer defects.
+The source file contains two frames (`SYSTEM4.BIN -> FORT.BIN`), while the first port rewrite contained
+four: the same gameplay pair plus `SAVE.BIN` and its nested helper. Native opcode `0x1ad` selects the
+terminal gameplay frame before opening modal helpers. Full restore through `0xae` reinstates that saved
+terminal context as the effective boundary. The managed restore cleared its temporary restore state but
+did not restore `_saveResumeFrame`, so the next `0x19e` fell back to the currently deepest frame. The port
+now marks the terminal restored `ExecFrame` before resuming at T1; a regression loads two frames, saves
+from a nested helper, and proves the rewrite still contains only the original two.
+
+The first fresh rewrite (`SAVE04`) then exposed the other half of that managed/native representation gap.
+It had two frames, but its restored SYSTEM4 ancestor changed from native `resume=8, call=8` to
+`resume=-1, call=-1`. Native `op_0xae_continue_save_load_stack_restore@0x416790` restores the parent
+context's real T1/T2/T3 coordinates before activating its child. The managed port instead leaves the
+parent physically parked at the synthetic `0xae` rendezvous while `RunFrame` recursively runs FORT.
+Recomputing a save record from that synthetic PC cannot find a T1 or T2 table entry. FORT itself loads
+and runs, but when stage launch unwinds through SYSTEM4, the missing T2 continuation enters ordinary
+Eushully boot.
+
+Each restored `ExecFrame` now shadows its original `NativeSavedScriptFrame` while a restored descendant
+is active. An intervening `0x19e` serializes those original T1/T2/T3 indices; T3 also reconstructs the
+live local-return stack, and the saved T1 coordinate seeds the frame's read-message state. The shadow is
+discarded only when that child genuinely returns and the parent resumes at its restored T2 continuation.
+Restored root-reload and process-exit outcomes also propagate through the synthetic recursion. A focused
+load -> nested save -> reload regression proves the parent indices survive and the child is not invoked
+again through ordinary `CallScript`.
+
+Manual slot 005 acceptance then proved that an immediate base rewrite could load and later launch a
+stage. Slot 006, authored immediately after entering that stage, still restored a black dungeon with
+visible but noninteractive UI. Its serialized FIELD range, 704 retained objects, map textures, and
+native-looking 0.8 transform were intact. A software replay reproduced the black result because FIELD
+had changed every restored map object's scale to `(0,0,1)`: global zoom selector `G[0x7682]` was 3, but
+the frame-local lookup table that should yield 80 was still zero.
+
+Ghidra resolves the ordering precisely. `script_frame_restore_saved_layout@0x40f2d0` calls
+`script_frame_load_resource@0x40e980`; that loader allocates and zeroes locals, loads the script body,
+and initializes its PC to the script codebase, not to `0xae`. The restored script therefore executes its
+ordinary entry prologue before reaching the rendezvous. FIELD's prologue constructs the inline zoom
+table `[40, 50, 64, 80, 100, 124, 156]`; the port's direct jump to `0xae` skipped those copies and let
+FIELD apply zero scale after the gfx record restore.
+
+Restored managed frames now likewise begin at script offset zero while retaining their saved T1/T2/T3
+coordinates for `0xae`. Synthetic root and child regressions prove both prologues execute. Replaying the
+unchanged real slot 006 now selects zoom 80, leaves the map at 0.8 scale, produces a nonblack dungeon
+raster, and reaches the gameplay poll. No slot rewrite is required for this correction.
+
+The same comparison exposed native retained-gfx pointer arithmetic. `context_state_serialize@0x40d320`
+writes a handle and copies the meaningful `0xb5` DWORD / `0x2d4`-byte record, then advances its DWORD
+cursor by `0x2d5`; `save_data_deserialize_and_begin_restore@0x40fd10` mirrors that stride. Early port
+output packed entries contiguously. In addition, `gfx_object_init_default@0x472810` seeds six identity
+matrices and three `0xffffffff` color defaults before the loader overwrites the complete record; the old
+writer began with zero bytes and only filled modeled fields. It also wrote translation vectors over the
+first row of the native 4x4 matrices instead of entries 12..14.
+
+`NativeNumberedSaveCodec` now emits the native `0xb54`-byte entry stride and recognizes the old tight
+experimental layout for compatibility. `NativeGfxPersistenceCodec` begins new records from the exact
+native defaults, writes row-vector translation matrices correctly, and retains the original raw record
+behind the semantic object so still-unnamed fields survive native load/save cycles. The `/v2` serializer
+and deserializer comments record the corrected stride and overwrite behavior.
+
 ### Opcode `0xae` continues numbered-save stack restoration (2026-07-20)
 
 Opcode `0xae` is the load-side rendezvous paired with serialized script-frame state. Its handler,
@@ -1816,6 +1883,12 @@ T1/T2/T3 indices back into live PC/call/return offsets. At the saved terminal co
 the restore flag and reinstates the saved context/return state. The corpus placement supports that control-flow role: 305 calls overwhelmingly follow
 coroutine-resume or call boundaries, including SC0000's main-loop resume sequence.
 
+The load order is significant: `script_frame_restore_saved_layout` delegates script creation to
+`script_frame_load_resource@0x40e980`, which initializes the new frame PC to the script codebase.
+Consequently the script runs its normal frame-local initialization and reaches `0xae` itself. Starting
+a restored frame directly at `0xae` is not native behavior; in FIELD it leaves the local zoom table
+zeroed and collapses the restored dungeon to a black point.
+
 The port now implements both branches: an ordinary one-instruction no-op outside restoration, and the
 T1/T2/T3-driven managed-frame reconstruction described above while a full numbered load is active.
 
@@ -2534,7 +2607,24 @@ backbuffer zero instead and record current target `-1` at manager `+0xb530` (`En
 `0x20e` then calls `d3d_clear_render_target_black@0x471460`, which invokes `IDirect3DDevice9::Clear` with no
 rectangles, `D3DCLEAR_TARGET | D3DCLEAR_ZBUFFER`, color zero, depth 1.0, and stencil zero. The port tracks
 the selected target and forwards the clear to the host; the retained compositor reconstructs its main
-backbuffer from black at publication boundaries, while offscreen clears also discard modeled text pixels.
+backbuffer from black at publication boundaries, while offscreen clears discard both modeled pixels and
+text draws.
+
+Opcode `0x222` publishes its retained handle range into whichever target `0x20d` selected; it is not
+merely an onscreen repaint request. SAVE.BIN provides a compact end-to-end example. It creates slot 2 at
+800x600, selects and clears it, then publishes handles `[0,0x130b0)` to capture the underlying gameplay
+without the save-menu objects. It next creates/selects/clears slot 192 at 112x84, binds slot 2 through
+handle `0x15f90`, scales that object to 14%, publishes the two-handle range, and passes slot 192 to
+`0x1ae`. Dungeon scripts use the same selected-target publication to construct map surfaces.
+
+The port originally retained only an object-list snapshot for transitions and requested an ordinary
+backbuffer repaint from `PresentObjectRange`; the selected surface's pixels remained its cleared black
+allocation. That produced completely black port-authored `.STH` files and caused restored dungeon map
+surfaces to turn black when their rebuild published offscreen. Godot now uses the platform-neutral
+software affine compositor for selected-target `0x222`/`0x20c` publication, including source surfaces,
+scaling, transforms, tint/opacity/blend, handle-range filtering, and real pixel clears. Backbuffer
+publication remains on the existing renderer. `gfx_present_object_range@0x482230` records the refined
+target ownership in the saved `/v2` image.
 
 This trace also corrects an important base-pointer assumption in earlier graphics notes. Offsets `+0x408`
 and `+0xb550` are relative to the retained-gfx manager at `EngineCtx+0x46614`, not to `EngineCtx` itself.

docs/opcode-reference.md

@@ -405,9 +405,9 @@ Implemented as a whole-stack root-reload boundary in the persistent VM. A reques
 ### 0xae `continue-save-load-stack-restore` (continue-save-load-stack-restore, argc 0)
 - **summary:** () - during serialized save restoration, replace the current frame PC with its saved resume/call target and advance through the saved script-context stack; otherwise a no-op.
 - **grounding:** source=investigation, confidence=high
-- **evidence:** Ghidra /v2: op_0xae_continue_save_load_stack_restore@0x416790 first tests ctx+0x53d24 (set by save_data_deserialize_and_begin_restore@0x40fd10). When clear it returns. When set, it selects the serialized frame layout through set:SaveVersion1/2, restores the current PC from that layout's saved return/call target, advances through contexts with script_frame_restore_saved_layout@0x40f2d0, and clears the restore flag on reaching the saved terminal context. Its 305 corpus sites overwhelmingly follow coroutine-resume/call boundaries, which provide the rendezvous points used while reconstructing the stack.
+- **evidence:** Ghidra /v2: op_0xae_continue_save_load_stack_restore@0x416790 first tests ctx+0x53d24 (set by save_data_deserialize_and_begin_restore@0x40fd10). When clear it returns. When set, it selects the serialized frame layout through set:SaveVersion1/2, restores the current PC from that layout's saved return/call target, advances through contexts with script_frame_restore_saved_layout@0x40f2d0, and clears the restore flag on reaching the saved terminal context. The restore helper calls script_frame_load_resource@0x40e980, which initializes each new frame PC to its script codebase; the script therefore executes its ordinary prologue before reaching 0xae. Native parent contexts retain restored coordinates while the child runs. SAVE00 -> SAVE03 proved the port must retain the terminal 0x1ad boundary; SAVE00 -> SAVE04 proved a managed parent parked at synthetic 0xae must serialize original resume=8/call=8 rather than synthetic -1/-1; slot 006 proved direct entry at 0xae skips FIELD's zoom-table prologue and collapses the dungeon map. Its 305 corpus sites overwhelmingly follow coroutine-resume/call boundaries.
 
-Layout 3 frame d259 indexes SYS4 T1 read-message reset sites, d260 indexes T2 call-script sites, and the saved local return stack indexes T3 local-call sites. Port status (2026-07-24): the active path reconstructs the saved recursive frame chain and resumes the terminal frame at its T1 boundary. The installed SAVE00 continuation gate proves SYSTEM4 -> FORT restoration reaches FORT's CHMENU gameplay poll; the synthetic gate asserts the child frame enters with SaveRestore rather than ordinary CallScript cause.
+Layout 3 frame d259 indexes SYS4 T1 read-message reset sites, d260 indexes T2 call-script sites, and the saved local return stack indexes T3 local-call sites. Port status (2026-07-24): each saved script is loaded at its ordinary entry, runs its frame-local prologue, and reaches 0xae itself; the active path then reconstructs the saved recursive frame chain, preserves each synthetic ancestor's original T1/T2/T3 coordinates while its restored child is active, reinstates the terminal restored frame as opcode 0x1ad's save boundary, and resumes it at T1. The installed SAVE00 continuation gate proves SYSTEM4 -> FORT restoration reaches FORT's CHMENU gameplay poll; the unchanged port-authored dungeon slot 006 proves FIELD's pre-rendezvous zoom table is initialized; synthetic gates assert both root/child prologue execution, SaveRestore rather than ordinary CallScript entry, exclusion of nested SAVE helpers from a re-save, and reload of that rewrite without ordinary boot re-entry.
 
 ### 0xc8 `sleep` (sleep, argc 1)
 - **summary:** Pause the current script for <duration> milliseconds while retained presentation continues.
@@ -790,9 +790,9 @@ Implemented through IHost.PlayModalMovieToSurface. Its operand uses the same nat
 - **evidence:** Ghidra 0x47ecc0 calls matrix builder 0x48afb1 for target obj+0x1ac. Consumer 0x472f00 uses delay obj+0x44, duration obj+0x58, current obj+0x16c, target obj+0x1ac, shared start obj+0x34, and retained-gfx frame time owner+0xb550 (EngineCtx+0x51b64).
 
 ### 0x222 `present-gfx-object-range` (present-gfx-object-range, argc 2)
-- **summary:** (first_handle)(count) - flush/present retained graphics objects in the selected handle range and clear their pending update flags.
+- **summary:** (first_handle)(count) - flush/present retained graphics objects in the selected handle range into the currently selected backbuffer or offscreen render target, then clear their pending update flags.
 - **grounding:** source=investigation, confidence=high
-- **evidence:** Ghidra /v2: op_0x222_handler@0x4235e0 calls gfx_present_object_range@0x482230. The worker enters the graphics service, walks the retained-object map, processes flagged objects whose handles fall in [first,first+count), clears pending flags, and finalizes the render batch. HISTORY.BIN uses (0,60000) after rebuilding its retained presentation.
+- **evidence:** Ghidra /v2: op_0x222_handler@0x4235e0 calls gfx_present_object_range@0x482230. The worker enters the graphics service, walks the retained-object map, processes flagged objects whose handles fall in [first,first+count), clears pending flags, and finalizes the render batch in the D3D target previously selected by op 0x20d. HISTORY.BIN uses (0,60000) for the backbuffer. SAVE.BIN instead renders [0,0x130b0) into 800x600 slot 2, then handle 0x15f90 at 14% scale into 112x84 slot 192; op 0x1ae writes slot 192 as the numbered .STH thumbnail.
 
 ### 0x223 `queue-surface-alpha-transition` (queue-surface-alpha-transition, argc 8)
 - **summary:** (command_key)(target_slot)(range_a_start)(range_a_count)(range_b_start)(range_b_count)(delay_ms)(duration_ms) — queue a type-0 timed alpha transition command in the separate ctx+0x414 command map. This is render-target/surface presentation state, not an object affine matrix. The render frame composites the two handle ranges into target_slot and ramps alpha 0->1 after delay over duration.
@@ -924,6 +924,13 @@ The setter get-or-creates the object and writes the complete operand. During ret
 - **grounding:** source=investigation, confidence=high
 - **evidence:** Ghidra /v2: op_0x24e_handler@0x425070 writes operand 1 directly to EngineCtx.gfx_animation_service_flags at +0x51b80. The mapped field is also read by op 0x243: bit 1 suppresses its force-complete/clock-reset request.
 
+### 0x258 `set-surface-persistence-flags` (set-surface-persistence-flags, argc 2)
+- **summary:** (surface_slot)(flags) - replace the native surface record's numbered-save persistence flags; bit 0 controls asset reload on restore and bit 1 controls the adjacent still-unnamed field.
+- **grounding:** source=investigation, confidence=high
+- **evidence:** Ghidra /v2: op_0x258_set_surface_persistence_flags@0x4250a0 reads the slot and flags operands and calls gfx_surface_set_persistence_flags@0x4159c0. The worker writes flags&1 to record +0x08 and (flags>>1)&1 to +0x0c in both 1,000-record tables. This corrects the old upstream address/name association: 0x422fe0 is opcode 0x20f movie playback, not 0x258. Corpus declaration chains follow opcode 0x259 at script entry.
+
+Scripts place declaration chains immediately after opcode 0x259 clears both fields. The port models bit 0 because it is consumed by layout-3 restoration; bit 1 is retained as an identified native field but has no known runtime consumer yet.
+
 ### 0x259 `script-entry` (script-entry, argc 0)
 - **summary:** zero-arg script/prologue entry; clears surface-record persistence fields +0x08 and +0x0c across both 1,000-record native tables before the declaration chain continues
 - **grounding:** source=investigation, confidence=high
@@ -1150,10 +1157,6 @@ Port status (2026-07-24): implemented through the same profile-lifetime setting
 - **summary:** 1 imm; mov->0x21b->stmt-end; near save/load-messkip — likely line/stmt id, verify not msg-control
 - **grounding:** source=harness, confidence=med, noop_headless=True
 
-### 0x258 `decl?` (u00422FE0, argc 2)
-- **summary:** 2 imm; runs in a chain right after script-entry 0x259, enumerating ids — prologue declaration/registration?
-- **grounding:** source=harness, confidence=low, noop_headless=True
-
 ## unknown
 
 ### 0x2 `exit` (exit, argc 0)

docs/phase-a-slice-plan.md

@@ -3628,6 +3628,111 @@ mapped and implemented as the script-entry clear of record `+0x08`/`+0x0c` (the
 Regressions cover native-default preservation, opt-in all-release, flag-for-flag port-authored output,
 script-entry clearing, and the installed all-zero reload-bit oracle.
 
+### Persistence implementation step 10 — publish selected offscreen render targets (2026-07-24)
+
+The first port-authored dungeon save exposed two symptoms with one graphics cause: its 112x84 `.STH`
+was completely black, and loading it showed the dungeon map for one frame before the map surface became
+black while the UI/minimap remained live.
+
+SAVE.BIN's bytecode makes the missing contract explicit. It renders the underlying gameplay handle range
+into created 800x600 surface 2, binds that surface to handle `0x15f90`, scales it to 14% into created
+112x84 surface 192, then asks opcode `0x1ae` to serialize surface 192. Both render passes use `0x20d`
+target selection, `0x20e` clear, and `0x222` range publication. Dungeon map construction uses the same
+offscreen path. Godot had modeled selection and black allocation, but `PresentObjectRange` only requested
+an onscreen recomposition, so neither target ever received pixels.
+
+Selected-target `0x222` and `0x20c` now publish through a platform-neutral retained-surface rasterizer.
+It filters the requested handle range and reproduces affine/range transforms, scaling, tint, opacity,
+blend mode, color keys, created/source surfaces, and black clears before storing the target pixels.
+Backbuffer publication is unchanged. A focused regression reproduces SAVE.BIN's scaled texture capture
+and proves objects outside the range cannot leak into the target. The original black SAVE01 thumbnail is
+a preserved diagnostic artifact; a newly overwritten save is required for visual acceptance of the
+fixed `.STH`. The subsequent immediate-resave test showed that a separate numbered-state round-trip defect
+still prevented the rewritten `.DAT` from resuming interactively.
+
+### Persistence implementation step 11 — repair port-authored numbered-state round trips (2026-07-24)
+
+The clean comparison target was `SAVE03`: load native slot 0 and save immediately, without entering a
+dungeon. Loading that rewrite was completely black, proving the remaining failure was not dungeon-map
+construction. The source slot has two saved frames; the rewrite had four and incorrectly included the
+currently open save-menu/helper frames. Full `0xae` restoration had not re-established the terminal saved
+frame as opcode `0x1ad`'s ongoing save cutoff, so the later `0x19e` fell back to the deepest active helper.
+The terminal restore now retains its `ExecFrame` as `_saveResumeFrame`. A regression reconstructs a
+two-frame save, invokes `0x19e` from a nested helper, and proves the new slot still contains only the two
+gameplay frames.
+
+Binary comparison also corrected the retained-gfx writer. Native records are not tightly packed:
+each handle plus meaningful `0x2d4`-byte record begins on a `0x2d5`-DWORD / `0xb54`-byte stride. The port
+now emits that sparse layout and imports its earlier packed experimental files. New records start from
+the exact `gfx_object_init_default` identity-matrix/color state; decoded native records retain an opaque
+byte template so unnamed fields survive rewrites; and translation is encoded at row-vector matrix
+entries 12..14 rather than over the matrix diagonal. Focused tests cover the native stride, all six
+identity matrices, translation placement, opaque-byte preservation, and the restored save boundary.
+
+The `/v2` serializer/deserializer annotations and canonical format notes are updated. The next acceptance
+action is to rebuild, load native slot 0, immediately overwrite a fresh slot, and load that fresh rewrite;
+older slot 1/2/3 files remain useful compatibility inputs but retain the already-written bad frame chain.
+
+Validation: all 391 engine tests pass; opcode lint/tooling tests are clean; the Godot C# build has zero
+warnings; the 211-object SAVE03 compatibility rewrite decodes with the native stride; and threaded
+headless execution reports `SELFTEST OK`.
+
+### Persistence implementation step 12 — preserve restored ancestor continuations (2026-07-24)
+
+Fresh slot 004 passed the immediate load gate, but selecting a stage played the Eushully intro. Its frame
+count was fixed at two, yet a decoded comparison found the decisive remaining difference:
+native slot 000's SYSTEM4 parent stores `resume=8, call=8`, while slot 004 stored `-1,-1`.
+
+Native `0xae` restores each parent context to its real T1/T2/T3 coordinates before activating the next
+saved child. The managed reconstruction instead recursively runs FORT while its SYSTEM4 `ExecFrame` is
+physically parked at the synthetic `0xae` instruction. Re-save capture derived coordinates from that
+synthetic PC, found neither a T1 nor T2 table entry, and wrote `-1`. FORT could initially run, but the
+later stage-launch unwind had no SYSTEM4 T2 continuation and fell into ordinary boot.
+
+Restored frames now shadow their original serialized coordinates while a restored descendant remains
+active. Re-saves use those T1/T2/T3 values; T3 rebuilds the live local-return stack; T1 initializes the
+read-message coordinate; and the shadow clears only when the child genuinely returns to its parent.
+Root-reload and process-exit results now propagate through this synthetic recursion as they do through
+ordinary `call-script`. The existing two-frame load/re-save regression now asserts the ancestor's resume
+and call indices and loads the rewrite again, rejecting any ordinary re-entry into the restored child.
+
+The `0xae` opcode source/reference, engine RE, roadmap, status memory, and `/v2` plate comment record the
+refinement. Existing slot 004 retains its already-written `-1,-1`; acceptance requires another fresh
+slot created from native slot 000, followed by load and stage launch.
+
+Validation: engine **392/392**, clean opcode lint/tooling, zero-warning Godot build, and threaded
+`SELFTEST OK`. The remaining gate is interactive confirmation with a newly authored slot; the fix cannot
+retroactively repair slot 004's serialized parent coordinates.
+
+### Persistence implementation step 13 — run restored script prologues before `0xae` (2026-07-24)
+
+Manual slot 005 acceptance passed both immediate base restoration and later stage launch. The next
+dungeon-authored rewrite, slot 006, loaded with its UI present but a black, noninteractive map. Binary
+and software-renderer inspection showed that the slot itself was healthy: it contained FIELD's saved
+range, 704 retained objects, map textures, and a native-looking 0.8 transform. During restore, however,
+FIELD overwrote the map objects with zero scale. Global zoom selector `G[0x7682]` was 3, while the
+frame-local zoom table entry that should have returned 80 was zero.
+
+Native `script_frame_restore_saved_layout@0x40f2d0` calls
+`script_frame_load_resource@0x40e980`, which creates each saved script frame at its ordinary codebase.
+The script runs its entry prologue and reaches `0xae` itself. The managed restore instead created frames
+directly at `0xae`, skipping FIELD's inline initialization of `[40, 50, 64, 80, 100, 124, 156]`.
+
+Restored frames now start at script offset zero while retaining the serialized T1/T2/T3 coordinates
+consumed at `0xae`. Synthetic root and child tests require their pre-rendezvous prologues to execute.
+The unchanged real slot 006 then replays with zoom 80, 0.8 object scale, a nonblack dungeon raster, and
+the gameplay poll; the fix does not require rewriting that save.
+
+The adjacent surface declaration gap is also closed. Native opcode `0x258` sets surface-record `+0x08`
+from flags bit 0 and `+0x0c` from bit 1 after `0x259` clears both tables. The port now applies bit 0 as
+the numbered-load reload policy; bit 1 remains identified but has no known consumer.
+
+Validation: engine **393/393**, clean opcode lint/tooling, zero-warning Godot build, threaded
+`SELFTEST OK`, and exact slot-006 software replay with zoom 80 and a nonblack dungeon raster.
+
+Acceptance: rebuild and interactively load existing slot 006. The expected result is the restored
+dungeon map and controls without a re-save.
+
 ## Data-semantics sidebar: focused append EBINIT inspection (2026-07-24)
 
 The static INIT surface now accepts a universal packed script id for focused append inspection.

docs/remake-architecture-and-roadmap.md

@@ -480,7 +480,19 @@ serialized mutable bank prefixes (preserving initialized unit/stage/string defin
 retained BGM/SFX state, preserves initialized flag-zero system surfaces while overlaying explicit saved
 reload records, reproduces the native configuration-gated all-surface release when explicitly enabled,
 and writes native per-slot reload/created metadata rather than treating every texture as reloadable.
-Opcode `0x259` supplies its real script-entry reload-policy clear. JSON inspection/export,
+Opcode `0x259` supplies its real script-entry reload-policy clear. Selected offscreen render targets now
+receive actual retained-object pixels, closing the black-thumbnail half of the first port-authored-save
+failure. The following immediate-resave oracle exposed and fixed two separate numbered-state defects:
+`0xae` now re-establishes the saved `0x1ad` gameplay-frame boundary before a later save UI opens, and
+retained objects use AGE's sparse stride plus native-initialized matrix/default bytes while preserving
+unnamed fields. The next fresh rewrite also proved that reconstructed managed ancestor frames must retain
+their original T1/T2/T3 coordinates while parked at synthetic `0xae`; preserving those coordinates fixes
+the later SYSTEM4 unwind that otherwise re-entered the Eushully intro. Slot 005 subsequently passed the
+base-load and stage-launch round trip. Dungeon-authored slot 006 exposed one further native-ordering
+requirement: restored scripts must begin at their ordinary entry, run frame-local prologues, and reach
+`0xae` themselves. Matching that order restores FIELD's 80% zoom table entry and produces the dungeon
+map from the unchanged slot; interactive slot-006 confirmation is the remaining visual gate.
+JSON inspection/export,
 namespaced mod data, and migrations remain additive extended-mode work rather than 1.0 compatibility
 requirements.
 

docs/sys4-format-notes.md

@@ -245,13 +245,23 @@ Retained graphics then uses:
 ```text
 0x00  gfx_record_size:u32              # 0x2d4
 +0x04  gfx_object_count:u32
-+0x08  repeated { handle:u32, record[0x2d4] }
++0x08  repeated sparse entries:
+         handle:u32
+         record[0x2d4]                 # 0xb5 meaningful DWORDs
+         padding[0x87c]
+       entry stride = 0x2d5 DWORDs = 0xb54 bytes
 ...    range_first:u32
 ...    range_count:i32
 ...    range_transform_record[0x2d4]
 ...    native allocation slack
 ```
 
+The sparse stride is not an allocator-only artifact. Both native writer and reader hold a DWORD pointer:
+after the handle, they copy `0x2d4` **bytes** but advance by `0x2d4` **DWORDs**. Consecutive handles are
+therefore `0x2d5` DWORDs apart. Early experimental port saves incorrectly packed
+`{handle,record}` contiguously; compatibility import recognizes that layout, while all new output uses
+the native sparse form.
+
 The 1,000 surface records preserve the native 20-byte metadata cells verbatim:
 
 | Offset | Size | Meaning |
@@ -270,9 +280,13 @@ surfaces expected to remain live across the load. The installed file is decisive
 are zero, while SYSTEM4's reusable choice-frame atlas remains recorded as resource `0x3383` in slot 15.
 A separate configuration-gated path can release all 1,000 surfaces before this loop, but Himegari's
 registered `CreateObject=1`, `AutoFreeTex=0` defaults leave it inactive. A retained `0x2d4`
-record is structurally complete but not every internal graphics field is semantically named. The native
+record is structurally complete but not every internal graphics field is semantically named. Native
-allocation is larger than the records actually written (`0x2e1 + object_count * 0x2d8` DWORDs in the
+`gfx_object_init_default` initializes six embedded 4x4 matrices (current/target scale, rotation, and
-graphics sizing term), leaving zero/slack bytes after the meaningful range record.
+translation) to identity, packed colors at `+0x60/+0x64/+0x240` to `0xffffffff`, and other fields to zero.
+Deserialization initializes an object and then overwrites all `0x2d4` bytes, so a compatibility writer
+must emit or preserve those defaults rather than zero-fill unnamed fields. The graphics sizing term remains
+larger than the meaningful sparse entries and range record
+(`0x2e1 + object_count * 0x2d8` DWORDs), leaving final allocation slack as well.
 
 The installed `SAVE00.DAT` validates the complete layout-3 decode: cutoff 1, global-bank counts
 `[402459,1,789,1,1,1]`, current BGM id `0x18`, retained SFX ids `0x3321` (channel 1) and

engine/Age.Engine.Tests/NativeNumberedSaveCodecTests.cs

@@ -28,6 +28,8 @@ public class NativeNumberedSaveCodecTests
                 new NativeSavedGfxObject(0xcf08,
                     Enumerable.Range(0, NativeNumberedSaveState.GfxRecordSize)
                         .Select(i => unchecked((byte)i)).ToArray()),
+                new NativeSavedGfxObject(0xcf09,
+                    Enumerable.Repeat((byte)0xa5, NativeNumberedSaveState.GfxRecordSize).ToArray()),
             ],
             RangeTransformFirst = 100,
             RangeTransformCount = 4,
@@ -53,7 +55,16 @@ public class NativeNumberedSaveCodecTests
         Assert.Equal(state.PointerGlobals, decoded.PointerGlobals);
         Assert.Equal(state.GfxObjects[0].Handle, decoded.GfxObjects[0].Handle);
         Assert.Equal(state.GfxObjects[0].Record, decoded.GfxObjects[0].Record);
+        Assert.Equal(state.GfxObjects[1].Handle, decoded.GfxObjects[1].Handle);
+        Assert.Equal(state.GfxObjects[1].Record, decoded.GfxObjects[1].Record);
+        Assert.False(decoded.LegacyTightGfxLayout);
         Assert.Equal(state.RangeTransformRecord, decoded.RangeTransformRecord);
+
+        const int nativeEntryStride = (1 + NativeNumberedSaveState.GfxRecordSize) * 4;
+        int objectsAt = FindGfxObjects(encoded, 2, 0xcf08);
+        Assert.Equal(0xcf08, BinaryPrimitives.ReadInt32LittleEndian(encoded.AsSpan(objectsAt)));
+        Assert.Equal(0xcf09, BinaryPrimitives.ReadInt32LittleEndian(
+            encoded.AsSpan(objectsAt + nativeEntryStride)));
     }
 
     [Fact]
@@ -153,4 +164,15 @@ public class NativeNumberedSaveCodecTests
                 BinaryPrimitives.ReadInt32LittleEndian(
                     state.SurfaceRecords.AsSpan(slot * 20 + 8))));
     }
+
+    private static int FindGfxObjects(byte[] payload, int count, int firstHandle)
+    {
+        for (int offset = 0; offset <= payload.Length - 12; offset += 4)
+            if (BinaryPrimitives.ReadInt32LittleEndian(payload.AsSpan(offset)) ==
+                    NativeNumberedSaveState.GfxRecordSize
+                && BinaryPrimitives.ReadInt32LittleEndian(payload.AsSpan(offset + 4)) == count
+                && BinaryPrimitives.ReadInt32LittleEndian(payload.AsSpan(offset + 8)) == firstHandle)
+                return offset + 8;
+        throw new Xunit.Sdk.XunitException("Encoded gfx table was not found.");
+    }
 }

engine/Age.Engine.Tests/NumberedSaveVmTests.cs

@@ -45,6 +45,8 @@ public class NumberedSaveVmTests
             vm.Gfx.SetSurfaceReloadOnRestore(7, true);
             vm.Gfx.ReleaseSurfaceRange(7, 1);
             vm.Gfx.BindDraw(100, 3, 1, 2, 30, 40, 50, 60);
+            vm.Gfx.SetCurrentTranslation(100, (7, 8, 9));
+            vm.Gfx.SetRotationChannel(100, 0, 500, (0, 0, 1), 90);
 
             vm.Run();
 
@@ -58,7 +60,25 @@ public class NumberedSaveVmTests
             Assert.Equal(0x3321, state.SoundEffectResourceIds[1]);
             Assert.Equal("姫狩り", state.StringGlobals[4]);
             Assert.Equal(0x77u, state.Frames.Single().ScriptId);
-            Assert.Contains(state.GfxObjects, item => item.Handle == 100);
+            NativeSavedGfxObject savedObject =
+                Assert.Single(state.GfxObjects, item => item.Handle == 100);
+            Assert.Equal(0, ReadGfxField(savedObject.Record, 0x34));
+            Assert.Equal(-1, ReadGfxField(savedObject.Record, 0x60));
+            Assert.Equal(0, ReadGfxField(savedObject.Record, 0x238));
+            Assert.Equal(-1, ReadGfxField(savedObject.Record, 0x240));
+            foreach (int matrixOffset in new[] { 0x6c, 0xac, 0xec, 0x16c, 0x1ac })
+            {
+                Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, matrixOffset));
+                Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, matrixOffset + 0x14));
+                Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, matrixOffset + 0x28));
+                Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, matrixOffset + 0x3c));
+            }
+            Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, 0x168));
+            Assert.Equal(BitConverter.SingleToInt32Bits(7), ReadGfxField(savedObject.Record, 0x19c));
+            Assert.Equal(BitConverter.SingleToInt32Bits(8), ReadGfxField(savedObject.Record, 0x1a0));
+            Assert.Equal(BitConverter.SingleToInt32Bits(9), ReadGfxField(savedObject.Record, 0x1a4));
+            Assert.Equal(BitConverter.SingleToInt32Bits(1), ReadGfxField(savedObject.Record, 0x130));
+            Assert.Equal(BitConverter.SingleToInt32Bits(-1), ReadGfxField(savedObject.Record, 0x13c));
             Assert.Equal(0x1234, ReadSurfaceField(state, 3, 0));
             Assert.Equal(0, ReadSurfaceField(state, 3, 8));
             Assert.Equal(-1, ReadSurfaceField(state, 4, 0));
@@ -88,14 +108,18 @@ public class NumberedSaveVmTests
             var store = new DirectoryNativeDatStore(root, Identity);
             Script resumed = WithTables(WithPackedId(ScriptAssembler.Assemble(Table, "RESUMED.BIN",
             [
+                (Table.ByLabel("mov")!.Value,
+                    [new Operand(GlobalInt, 0x510), new Operand(Immediate, 11)]),
                 (0xae, Array.Empty<Operand>()),
                 (0x3, [new Operand(Immediate, 0x89)]),
                 (Table.ByLabel("mov")!.Value,
                     [new Operand(GlobalInt, 0x500), new Operand(GlobalInt, 0x123)]),
                 (0x2, Array.Empty<Operand>()),
-            ], []), 0x88), scriptCallOffsets: [1]);
+            ], []), 0x88), scriptCallOffsets: [6]);
             Script child = WithPackedId(ScriptAssembler.Assemble(Table, "CHILD.BIN",
             [
+                (Table.ByLabel("mov")!.Value,
+                    [new Operand(GlobalInt, 0x511), new Operand(Immediate, 22)]),
                 (0xae, Array.Empty<Operand>()),
                 (Table.ByLabel("mov")!.Value,
                     [new Operand(GlobalInt, 0x501), new Operand(GlobalInt, 0x123)]),
@@ -168,6 +192,8 @@ public class NumberedSaveVmTests
             Assert.Equal(777, vm.Globals[0x124]);
             Assert.Equal(456, vm.Globals[0x500]);
             Assert.Equal(456, vm.Globals[0x501]);
+            Assert.Equal(11, vm.Globals[0x510]);
+            Assert.Equal(22, vm.Globals[0x511]);
             Assert.Equal(1, vm.Globals[0x502]);
             Assert.Equal("復帰", vm.GlobalStrings[0]);
             Assert.Equal(0x123, vm.GlobalPointers[0]);
@@ -204,6 +230,163 @@ public class NumberedSaveVmTests
         }
     }
 
+    [Fact]
+    public void FullLoadReestablishesSaveBoundaryBeforeNestedSaveUiWritesAnotherSlot()
+    {
+        string root = NewTemporaryDirectory();
+        try
+        {
+            var store = new DirectoryNativeDatStore(root, Identity);
+            int callScript = Table.ByLabel("call-script")!.Value;
+            Script resumed = WithTables(WithPackedId(ScriptAssembler.Assemble(Table, "RESUMED.BIN",
+            [
+                (0xae, Array.Empty<Operand>()),
+                (callScript, [new Operand(Immediate, 0x89)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x88), scriptCallOffsets: [1]);
+            Script child = WithPackedId(ScriptAssembler.Assemble(Table, "CHILD.BIN",
+            [
+                (0xae, Array.Empty<Operand>()),
+                (callScript, [new Operand(Immediate, 0x91)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x89);
+            Script saveUi = WithPackedId(ScriptAssembler.Assemble(Table, "SAVE_UI.BIN",
+            [
+                (0x19e, [new Operand(LocalInt, 0), new Operand(Immediate, 2)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x91);
+            Script loader = WithPackedId(ScriptAssembler.Assemble(Table, "LOADER.BIN",
+            [
+                (0x1a1, [new Operand(LocalInt, 0), new Operand(Immediate, 1)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x99);
+            byte[] opaqueRecord = NativeGfxRecord(3);
+            BinaryPrimitives.WriteInt32LittleEndian(opaqueRecord.AsSpan(0x68), 0x12345678);
+            NativeNumberedSaveState source = NativeNumberedSaveCodec.Empty(
+            [
+                new NativeSavedScriptFrame(-1, 0x88, [], 8, 0),
+                new NativeSavedScriptFrame(0, 0x89, [], -1, -1),
+            ]) with
+            {
+                GfxObjects = [new NativeSavedGfxObject(100, opaqueRecord)],
+            };
+            store.SaveNumberedFile(
+                1, NativeNumberedSaveCodec.Encode(source), [],
+                NativeSystemTime.FromLocalDateTime(DateTime.Now), 0);
+            var vm = new VirtualMachine(
+                loader, Table, new RecordingHost(),
+                provider: new MapProvider(new()
+                {
+                    [0x88] = resumed,
+                    [0x89] = child,
+                    [0x91] = saveUi,
+                }),
+                nativeDatStore: store);
+
+            vm.Run();
+
+            NativeNumberedSaveFile? resaved = store.LoadNumberedFile(2);
+            Assert.NotNull(resaved);
+            NativeNumberedSaveState state =
+                NativeNumberedSaveCodec.Decode(resaved!.Document.Payload);
+            Assert.Equal([0x88u, 0x89u], state.Frames.Select(frame => frame.ScriptId));
+            Assert.Equal(8, state.Frames[0].ResumeIndex);
+            Assert.Equal(0, state.Frames[0].CallTargetIndex);
+            Assert.Equal(
+                0x12345678,
+                ReadGfxField(Assert.Single(state.GfxObjects).Record, 0x68));
+
+            Script reloader = WithPackedId(ScriptAssembler.Assemble(Table, "RELOADER.BIN",
+            [
+                (0x1a1, [new Operand(LocalInt, 0), new Operand(Immediate, 2)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x9a);
+            var trace = new RecordingTraceSink();
+            var roundTrippedVm = new VirtualMachine(
+                reloader, Table, new RecordingHost(),
+                provider: new MapProvider(new()
+                {
+                    [0x88] = resumed,
+                    [0x89] = child,
+                    [0x91] = saveUi,
+                }),
+                sink: trace,
+                nativeDatStore: store);
+
+            roundTrippedVm.Run();
+
+            Assert.DoesNotContain(trace.Events, item =>
+                item.Kind == Age.Engine.Diagnostics.TraceEventKind.FrameEnter
+                && item.Name == "CHILD.BIN"
+                && item.Cause == Age.Engine.Diagnostics.FrameCause.CallScript);
+            Assert.Equal("exit", roundTrippedVm.HaltReason);
+        }
+        finally
+        {
+            Directory.Delete(root, recursive: true);
+        }
+    }
+
+    [Fact]
+    public void RestoredChildRootReloadDiscardsSyntheticAncestorChain()
+    {
+        string root = NewTemporaryDirectory();
+        try
+        {
+            var store = new DirectoryNativeDatStore(root, Identity);
+            int callScript = Table.ByLabel("call-script")!.Value;
+            int move = Table.ByLabel("mov")!.Value;
+            Script resumed = WithTables(WithPackedId(ScriptAssembler.Assemble(Table, "RESUMED.BIN",
+            [
+                (0xae, Array.Empty<Operand>()),
+                (callScript, [new Operand(Immediate, 0x89)]),
+                (move, [new Operand(GlobalInt, 0x600), new Operand(Immediate, 1)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x88), scriptCallOffsets: [1]);
+            Script child = WithPackedId(ScriptAssembler.Assemble(Table, "CHILD_RELOAD.BIN",
+            [
+                (0xae, Array.Empty<Operand>()),
+                (0x9, Array.Empty<Operand>()),
+            ], []), 0x89);
+            Script reloaded = WithPackedId(ScriptAssembler.Assemble(Table, "RELOADED.BIN",
+            [
+                (move, [new Operand(GlobalInt, 0x601), new Operand(Immediate, 1)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0);
+            Script loader = WithPackedId(ScriptAssembler.Assemble(Table, "LOADER.BIN",
+            [
+                (0x1a1, [new Operand(LocalInt, 0), new Operand(Immediate, 1)]),
+                (0x2, Array.Empty<Operand>()),
+            ], []), 0x99);
+            NativeNumberedSaveState source = NativeNumberedSaveCodec.Empty(
+            [
+                new NativeSavedScriptFrame(-1, 0x88, [], -1, 0),
+                new NativeSavedScriptFrame(0, 0x89, [], -1, -1),
+            ]);
+            store.SaveNumberedFile(
+                1, NativeNumberedSaveCodec.Encode(source), [],
+                NativeSystemTime.FromLocalDateTime(DateTime.Now), 0);
+            var vm = new VirtualMachine(
+                loader, Table, new RecordingHost(),
+                provider: new MapProvider(new()
+                {
+                    [0] = reloaded,
+                    [0x88] = resumed,
+                    [0x89] = child,
+                }),
+                nativeDatStore: store);
+
+            vm.Run();
+
+            Assert.False(vm.Globals.ContainsKey(0x600));
+            Assert.Equal(1, vm.Globals[0x601]);
+        }
+        finally
+        {
+            Directory.Delete(root, recursive: true);
+        }
+    }
+
     [Fact]
     public void DataOnlyLoadReleasesAllSurfacesWhenBothNativeSettingsEnableIt()
     {
@@ -272,6 +455,28 @@ public class NumberedSaveVmTests
         Assert.False(surface.ReloadOnRestore);
     }
 
+    [Fact]
+    public void SurfacePersistenceDeclarationSetsAndClearsReloadBit()
+    {
+        Script script = ScriptAssembler.Assemble(Table, "SURFACE_POLICY.BIN",
+        [
+            (0x258, [new Operand(Immediate, 15), new Operand(Immediate, 1)]),
+            (0x258, [new Operand(Immediate, 16), new Operand(Immediate, 3)]),
+            (0x258, [new Operand(Immediate, 15), new Operand(Immediate, 0)]),
+            (0x2, Array.Empty<Operand>()),
+        ], []);
+        var vm = new VirtualMachine(script, Table, new RecordingHost());
+        vm.Gfx.SetSurface(15, 0x3383, 0);
+        vm.Gfx.SetSurface(16, 0x3384, 0);
+
+        vm.Run();
+
+        GfxSurfacePersistenceState[] surfaces =
+            vm.Gfx.CapturePersistenceSnapshot().Surfaces.ToArray();
+        Assert.False(Assert.Single(surfaces, item => item.Slot == 15).ReloadOnRestore);
+        Assert.True(Assert.Single(surfaces, item => item.Slot == 16).ReloadOnRestore);
+    }
+
     private static Script WithPackedId(Script source, uint packedId)
         => new()
         {
@@ -342,11 +547,21 @@ public class NumberedSaveVmTests
         Write(0x24, 50);
         Write(0x28, 60);
         Write(0x60, -1);
-        Write(0x238, 1);
+        Write(0x64, -1);
-        Write(0x23c, 1);
+        foreach (int matrixOffset in new[] { 0x6c, 0xac, 0xec, 0x12c, 0x16c, 0x1ac })
+        {
+            Write(matrixOffset, BitConverter.SingleToInt32Bits(1));
+            Write(matrixOffset + 0x14, BitConverter.SingleToInt32Bits(1));
+            Write(matrixOffset + 0x28, BitConverter.SingleToInt32Bits(1));
+            Write(matrixOffset + 0x3c, BitConverter.SingleToInt32Bits(1));
+        }
+        Write(0x240, -1);
         return result;
     }
 
+    private static int ReadGfxField(byte[] record, int offset)
+        => BinaryPrimitives.ReadInt32LittleEndian(record.AsSpan(offset));
+
     private static string NewTemporaryDirectory()
     {
         string path = Path.Combine(Path.GetTempPath(), "age-save-vm-" + Guid.NewGuid().ToString("N"));

engine/Age.Engine.Tests/RetainedSurfaceRasterizerTests.cs

@@ -0,0 +1,41 @@
+using Age.Engine.Model;
+using Age.Engine.Sys4;
+
+public class RetainedSurfaceRasterizerTests
+{
+    [Fact]
+    public void CompositeRangePublishesScaledRetainedTextureIntoOffscreenSurface()
+    {
+        var gfx = new GfxState();
+        gfx.SetSurface(2, 0x1234, -1);
+        gfx.BindDraw(100, 2, 0, 0, 4, 4, 0, 0);
+        gfx.SetCurrentScale(100, (50, 50, 100));
+        gfx.SetSurface(3, 0x5678, -1);
+        gfx.BindDraw(200, 3, 0, 0, 2, 2, 0, 0);
+
+        byte[] sourcePixels = new byte[4 * 4 * 4];
+        for (int index = 0; index < sourcePixels.Length; index += 4)
+        {
+            sourcePixels[index] = 0x44;
+            sourcePixels[index + 1] = 0x88;
+            sourcePixels[index + 2] = 0xcc;
+            sourcePixels[index + 3] = 0xff;
+        }
+        var source = new RgbaImage(4, 4, sourcePixels);
+        var excluded = new RgbaImage(2, 2, Enumerable.Repeat((byte)0xff, 16).ToArray());
+        var destination = new RgbaImage(2, 2, new byte[16]);
+
+        int rendered = RetainedSurfaceRasterizer.CompositeRange(
+            destination, gfx.SnapshotVisibleObjects(), 100, 1,
+            item => item.Handle == 100 ? source : excluded);
+
+        Assert.Equal(1, rendered);
+        for (int index = 0; index < destination.Pixels.Length; index += 4)
+        {
+            Assert.Equal(0x44, destination.Pixels[index]);
+            Assert.Equal(0x88, destination.Pixels[index + 1]);
+            Assert.Equal(0xcc, destination.Pixels[index + 2]);
+            Assert.Equal(0xff, destination.Pixels[index + 3]);
+        }
+    }
+}

engine/Age.Engine/Model/GfxState.cs

@@ -101,6 +101,10 @@ public sealed class GfxState
     }
     public sealed class GfxObject
     {
+        // The native numbered-save record contains several full transform matrices and reserved fields
+        // beyond the port's semantic model. Preserve the decoded bytes so a native load/save round trip
+        // does not erase state that the port has not modeled yet; known fields are patched by the codec.
+        public byte[]? NativePersistenceRecord;
         public (long X, long Y, long Z) V18, V24, V16c;
         public long Field64, Field68, Field6c;
         public long Color = 0xffffffff; // native gfx_object_init_default obj+0x60: identity packed ARGB
@@ -464,6 +468,7 @@ public sealed class GfxState
     private static GfxObject CloneState(GfxObject s)
         => new()
         {
+            NativePersistenceRecord = s.NativePersistenceRecord?.ToArray(),
             V18 = s.V18, V24 = s.V24, V16c = s.V16c,
             Field64 = s.Field64, Field68 = s.Field68, Field6c = s.Field6c,
             Color = s.Color, HasColor = s.HasColor, StaticColorMode = s.StaticColorMode,

engine/Age.Engine/Persistence/NativeGfxPersistenceCodec.cs

@@ -55,18 +55,27 @@ internal static class NativeGfxPersistenceCodec
                     reload));
         }
         var objects = state.GfxObjects
-            .Select(item => (item.Handle, DecodeObject(item.Record)))
+            .Select(item => (item.Handle, DecodeObject(item.Record, state.LegacyTightGfxLayout)))
             .ToArray();
         return new GfxPersistenceSnapshot(
             surfaces, objects, state.RangeTransformFirst, state.RangeTransformCount,
-            DecodeObject(state.RangeTransformRecord));
+            DecodeObject(state.RangeTransformRecord, state.LegacyTightGfxLayout));
     }
 
     private static byte[] EncodeObject(GfxState.GfxObject value)
     {
-        byte[] raw = new byte[NativeNumberedSaveState.GfxRecordSize];
+        byte[] raw = value.NativePersistenceRecord is { Length: NativeNumberedSaveState.GfxRecordSize }
-        int flags = value.Visible ? 1 : 0;
+            ? value.NativePersistenceRecord.ToArray()
-        if (value.RotationEnabled || value.SrcAnim || value.ColorAnim) flags |= 4;
+            : CreateDefaultObjectRecord();
+        int flags = ReadInt(raw, 0);
+        flags = value.Visible ? flags | 1 : flags & ~1;
+        flags = value.OneShotColorEnabled || value.ScaleEnabled
+            || value.RotationChannelEnabled || value.TranslationEnabled
+            ? flags | 2
+            : flags & ~2;
+        flags = value.RotationEnabled || value.SrcAnim || value.ColorAnim
+            ? flags | 4
+            : flags & ~4;
         WriteInt(raw, 0, flags);
         WriteInt(raw, 4, value.SourceSlot);
         WriteInt(raw, 8, value.SrcRect.X);
@@ -76,7 +85,7 @@ internal static class NativeGfxPersistenceCodec
         WriteVector(raw, 0x18, value.V18);
         WriteVector(raw, 0x24, value.V24);
         WriteInt(raw, 0x30, unchecked((int)value.StaticColorMode));
-        WriteInt(raw, 0x34, unchecked((int)value.OneShotStartMs));
+        WriteInt(raw, 0x34, EncodeNativeStart(value.OneShotStartMs));
         WriteInt(raw, 0x38, unchecked((int)value.ColorDelayMs));
         WriteInt(raw, 0x3c, unchecked((int)value.ScaleDelayMs));
         WriteInt(raw, 0x40, unchecked((int)value.RotationDelayMs));
@@ -89,27 +98,34 @@ internal static class NativeGfxPersistenceCodec
         WriteInt(raw, 0x64, unchecked((int)value.OneShotColorTarget));
         WriteScaleMatrix(raw, 0x6c, value.ScaleCurrent);
         WriteScaleMatrix(raw, 0xac, value.ScaleTarget);
-        WriteFloat(raw, 0x16c, value.TranslationCurrent.X);
+        WriteRotationMatrix(raw, 0xec, value.RotationCurrent);
-        WriteFloat(raw, 0x170, value.TranslationCurrent.Y);
+        WriteRotationMatrix(raw, 0x12c, value.RotationTarget);
-        WriteFloat(raw, 0x174, value.TranslationCurrent.Z);
+        WriteTranslationMatrix(raw, 0x16c, value.TranslationCurrent);
-        WriteFloat(raw, 0x1ac, value.TranslationTarget.X);
+        WriteTranslationMatrix(raw, 0x1ac, value.TranslationTarget);
-        WriteFloat(raw, 0x1b0, value.TranslationTarget.Y);
+        WriteFloat(raw, 0x1ec, value.RotationCurrent.X);
-        WriteFloat(raw, 0x1b4, value.TranslationTarget.Z);
+        WriteFloat(raw, 0x1f0, value.RotationCurrent.Y);
-        WriteInt(raw, 0x20c, unchecked((int)value.ColorStart));
+        WriteFloat(raw, 0x1f4, value.RotationCurrent.Z);
-        WriteInt(raw, 0x214, unchecked((int)value.RotationStartMs));
+        WriteFloat(raw, 0x1f8, value.RotationTarget.X);
+        WriteFloat(raw, 0x1fc, value.RotationTarget.Y);
+        WriteFloat(raw, 0x200, value.RotationTarget.Z);
+        WriteFloat(raw, 0x204, value.RotationCurrent.Angle);
+        WriteFloat(raw, 0x208, value.RotationTarget.Angle);
+        WriteInt(raw, 0x20c, EncodeNativeStart(value.ColorStart));
+        WriteInt(raw, 0x214, EncodeNativeStart(value.RotationStartMs));
         WriteInt(raw, 0x220, unchecked((int)value.ColorPeriod));
         WriteInt(raw, 0x228, unchecked((int)value.RotationPeriodMs));
         WriteInt(raw, 0x230, unchecked((int)value.SrcPeriod));
         WriteInt(raw, 0x234, unchecked((int)value.SrcCell));
-        WriteInt(raw, 0x238, unchecked((int)value.SrcFrameCount));
+        WriteInt(raw, 0x238, value.SrcAnim ? unchecked((int)value.SrcFrameCount) : 0);
-        WriteInt(raw, 0x23c, unchecked((int)value.SrcColumns));
+        WriteInt(raw, 0x23c, value.SrcAnim ? unchecked((int)value.SrcColumns) : 0);
-        WriteInt(raw, 0x240, unchecked((int)value.ColorTarget));
+        if (value.ColorAnim)
+            WriteInt(raw, 0x240, unchecked((int)value.ColorTarget));
         WriteVector(raw, 0x244, value.RotationAxis);
         WriteInt(raw, 0x2d0, unchecked((int)value.OneShotAnimationControlFlags));
         return raw;
     }
 
-    private static GfxState.GfxObject DecodeObject(ReadOnlySpan<byte> raw)
+    private static GfxState.GfxObject DecodeObject(ReadOnlySpan<byte> raw, bool legacyTightLayout)
     {
         if (raw.Length != NativeNumberedSaveState.GfxRecordSize)
             throw new InvalidDataException("Native retained-gfx record has the wrong size.");
@@ -118,13 +134,16 @@ internal static class NativeGfxPersistenceCodec
         int flags = ReadInt(raw, 0);
         return new GfxState.GfxObject
         {
+            NativePersistenceRecord = legacyTightLayout
+                ? CreateDefaultObjectRecord()
+                : raw.ToArray(),
             Visible = (flags & 1) != 0,
             SourceSlot = ReadInt(raw, 4),
             SrcRect = (left, top, right - left, bottom - top),
             V18 = ReadLongVector(raw, 0x18),
             V24 = ReadLongVector(raw, 0x24),
             StaticColorMode = ReadInt(raw, 0x30),
-            OneShotStartMs = ReadInt(raw, 0x34),
+            OneShotStartMs = DecodeNativeStart(ReadInt(raw, 0x34)),
             ColorDelayMs = ReadInt(raw, 0x38),
             ScaleDelayMs = ReadInt(raw, 0x3c),
             RotationDelayMs = ReadInt(raw, 0x40),
@@ -138,11 +157,17 @@ internal static class NativeGfxPersistenceCodec
             OneShotColorTarget = unchecked((uint)ReadInt(raw, 0x64)),
             ScaleCurrent = ReadScale(raw, 0x6c),
             ScaleTarget = ReadScale(raw, 0xac),
-            TranslationCurrent = ReadDoubleVector(raw, 0x16c),
+            TranslationCurrent = ReadMatrixTranslation(raw, 0x16c),
-            V16c = ReadLongFloatVector(raw, 0x16c),
+            V16c = ReadLongMatrixTranslation(raw, 0x16c),
-            TranslationTarget = ReadDoubleVector(raw, 0x1ac),
+            TranslationTarget = ReadMatrixTranslation(raw, 0x1ac),
-            ColorStart = ReadInt(raw, 0x20c),
+            RotationCurrent = (
-            RotationStartMs = ReadInt(raw, 0x214),
+                ReadFloat(raw, 0x1ec), ReadFloat(raw, 0x1f0), ReadFloat(raw, 0x1f4),
+                ReadFloat(raw, 0x204)),
+            RotationTarget = (
+                ReadFloat(raw, 0x1f8), ReadFloat(raw, 0x1fc), ReadFloat(raw, 0x200),
+                ReadFloat(raw, 0x208)),
+            ColorStart = DecodeNativeStart(ReadInt(raw, 0x20c)),
+            RotationStartMs = DecodeNativeStart(ReadInt(raw, 0x214)),
             ColorPeriod = ReadInt(raw, 0x220),
             RotationPeriodMs = ReadInt(raw, 0x228),
             SrcPeriod = ReadInt(raw, 0x230),
@@ -170,6 +195,7 @@ internal static class NativeGfxPersistenceCodec
 
     private static void WriteScaleMatrix(Span<byte> raw, int offset, (double X, double Y, double Z) scale)
     {
+        ClearMatrix(raw, offset);
         WriteFloat(raw, offset, scale.X);
         WriteFloat(raw, offset + 0x14, scale.Y);
         WriteFloat(raw, offset + 0x28, scale.Z);
@@ -179,6 +205,40 @@ internal static class NativeGfxPersistenceCodec
     private static (double X, double Y, double Z) ReadScale(ReadOnlySpan<byte> raw, int offset)
         => (ReadFloat(raw, offset), ReadFloat(raw, offset + 0x14), ReadFloat(raw, offset + 0x28));
 
+    private static void WriteRotationMatrix(
+        Span<byte> raw, int offset, (double X, double Y, double Z, double Angle) rotation)
+    {
+        ClearMatrix(raw, offset);
+        double length = Math.Sqrt(
+            rotation.X * rotation.X + rotation.Y * rotation.Y + rotation.Z * rotation.Z);
+        if (length <= double.Epsilon || rotation.Angle == 0)
+        {
+            WriteFloat(raw, offset, 1);
+            WriteFloat(raw, offset + 0x14, 1);
+            WriteFloat(raw, offset + 0x28, 1);
+            WriteFloat(raw, offset + 0x3c, 1);
+            return;
+        }
+
+        double x = rotation.X / length;
+        double y = rotation.Y / length;
+        double z = rotation.Z / length;
+        double radians = rotation.Angle * Math.PI / 180.0;
+        double cosine = Math.Cos(radians);
+        double sine = Math.Sin(radians);
+        double complement = 1.0 - cosine;
+        WriteFloat(raw, offset, x * x * complement + cosine);
+        WriteFloat(raw, offset + 0x04, x * y * complement + z * sine);
+        WriteFloat(raw, offset + 0x08, x * z * complement - y * sine);
+        WriteFloat(raw, offset + 0x10, x * y * complement - z * sine);
+        WriteFloat(raw, offset + 0x14, y * y * complement + cosine);
+        WriteFloat(raw, offset + 0x18, x * sine + y * z * complement);
+        WriteFloat(raw, offset + 0x20, y * sine + x * z * complement);
+        WriteFloat(raw, offset + 0x24, y * z * complement - x * sine);
+        WriteFloat(raw, offset + 0x28, z * z * complement + cosine);
+        WriteFloat(raw, offset + 0x3c, 1);
+    }
+
     private static void WriteVector(Span<byte> raw, int offset, (long X, long Y, long Z) vector)
     {
         WriteInt(raw, offset, unchecked((int)vector.X));
@@ -189,11 +249,55 @@ internal static class NativeGfxPersistenceCodec
     private static (long X, long Y, long Z) ReadLongVector(ReadOnlySpan<byte> raw, int offset)
         => (ReadInt(raw, offset), ReadInt(raw, offset + 4), ReadInt(raw, offset + 8));
 
-    private static (long X, long Y, long Z) ReadLongFloatVector(ReadOnlySpan<byte> raw, int offset)
+    private static void WriteTranslationMatrix(
-        => ((long)ReadFloat(raw, offset), (long)ReadFloat(raw, offset + 4), (long)ReadFloat(raw, offset + 8));
+        Span<byte> raw, int offset, (double X, double Y, double Z) translation)
+    {
+        ClearMatrix(raw, offset);
+        WriteFloat(raw, offset, 1);
+        WriteFloat(raw, offset + 0x14, 1);
+        WriteFloat(raw, offset + 0x28, 1);
+        WriteFloat(raw, offset + 0x30, translation.X);
+        WriteFloat(raw, offset + 0x34, translation.Y);
+        WriteFloat(raw, offset + 0x38, translation.Z);
+        WriteFloat(raw, offset + 0x3c, 1);
+    }
 
-    private static (double X, double Y, double Z) ReadDoubleVector(ReadOnlySpan<byte> raw, int offset)
+    private static (long X, long Y, long Z) ReadLongMatrixTranslation(
-        => (ReadFloat(raw, offset), ReadFloat(raw, offset + 4), ReadFloat(raw, offset + 8));
+        ReadOnlySpan<byte> raw, int offset)
+        => ((long)ReadFloat(raw, offset + 0x30),
+            (long)ReadFloat(raw, offset + 0x34),
+            (long)ReadFloat(raw, offset + 0x38));
+
+    private static (double X, double Y, double Z) ReadMatrixTranslation(
+        ReadOnlySpan<byte> raw, int offset)
+        => (ReadFloat(raw, offset + 0x30),
+            ReadFloat(raw, offset + 0x34),
+            ReadFloat(raw, offset + 0x38));
+
+    private static void ClearMatrix(Span<byte> raw, int offset)
+        => raw.Slice(offset, 0x40).Clear();
+
+    private static int EncodeNativeStart(long value)
+        => value < 0 ? 0 : unchecked((int)value);
+
+    private static long DecodeNativeStart(int value)
+        => value == 0 ? -1 : value;
+
+    private static byte[] CreateDefaultObjectRecord()
+    {
+        byte[] raw = new byte[NativeNumberedSaveState.GfxRecordSize];
+        WriteInt(raw, 0x60, -1);
+        WriteInt(raw, 0x64, -1);
+        foreach (int matrixOffset in new[] { 0x6c, 0xac, 0xec, 0x12c, 0x16c, 0x1ac })
+        {
+            WriteFloat(raw, matrixOffset, 1);
+            WriteFloat(raw, matrixOffset + 0x14, 1);
+            WriteFloat(raw, matrixOffset + 0x28, 1);
+            WriteFloat(raw, matrixOffset + 0x3c, 1);
+        }
+        WriteInt(raw, 0x240, -1);
+        return raw;
+    }
 
     private static void WriteFloat(Span<byte> raw, int offset, double value)
         => WriteInt(raw, offset, BitConverter.SingleToInt32Bits((float)value));

engine/Age.Engine/Persistence/NativeNumberedSaveCodec.cs

@@ -28,7 +28,8 @@ public sealed record NativeNumberedSaveState(
     IReadOnlyList<NativeSavedGfxObject> GfxObjects,
     long RangeTransformFirst,
     int RangeTransformCount,
-    byte[] RangeTransformRecord)
+    byte[] RangeTransformRecord,
+    bool LegacyTightGfxLayout = false)
 {
     public const int SoundEffectChannelCount = 10;
     public const int ResourceRecordsSize = 300 * 4;
@@ -45,6 +46,9 @@ public static class NativeNumberedSaveCodec
     private const int FrameReturnCapacity = 256;
     private const int GfxAllocationDwords = 0x2d8;
     private const int GfxConstantDwords = 0x2e1;
+    // AGE stores a byte-sized 0x2d4 object record at the front of a 0x2d4-DWORD region.
+    // The handle consumes one preceding DWORD, so consecutive entries start 0x2d5 DWORDs apart.
+    private const int GfxEntryStrideBytes = (1 + NativeNumberedSaveState.GfxRecordSize) * 4;
     private static readonly Encoding NativeEncoding = CreateNativeEncoding();
 
     public static byte[] Encode(NativeNumberedSaveState state)
@@ -103,7 +107,7 @@ public static class NativeNumberedSaveCodec
         {
             WriteInt(payload, at, unchecked((int)gfx.Handle));
             gfx.Record.CopyTo(payload, at + 4);
-            at += 4 + NativeNumberedSaveState.GfxRecordSize;
+            at += GfxEntryStrideBytes;
         }
         WriteInt(payload, at, unchecked((int)state.RangeTransformFirst));
         WriteInt(payload, at + 4, state.RangeTransformCount);
@@ -154,14 +158,19 @@ public static class NativeNumberedSaveCodec
         if (gfxRecordSize != NativeNumberedSaveState.GfxRecordSize)
             throw new InvalidDataException($"Unsupported native gfx record size 0x{gfxRecordSize:x}.");
         at += 8;
-        var objects = new NativeSavedGfxObject[gfxCount];
+        int objectsAt = at;
-        for (int i = 0; i < objects.Length; i++)
+        (NativeSavedGfxObject[] objects, int afterObjects) =
+            ReadGfxObjects(payload, objectsAt, gfxCount, GfxEntryStrideBytes);
+        bool legacyTightGfxLayout = false;
+        // Compatibility for experimental files written by the port before the native DWORD stride was
+        // understood. A native retained-object map cannot contain duplicate handles.
+        if (objects.Select(item => item.Handle).Distinct().Count() != objects.Length)
         {
-            Require(payload, at, 4 + gfxRecordSize, "numbered-save gfx object");
+            (objects, afterObjects) =
-            long handle = ReadInt(payload, at);
+                ReadGfxObjects(payload, objectsAt, gfxCount, 4 + gfxRecordSize);
-            objects[i] = new NativeSavedGfxObject(handle, payload.Slice(at + 4, gfxRecordSize).ToArray());
+            legacyTightGfxLayout = true;
-            at += 4 + gfxRecordSize;
         }
+        at = afterObjects;
         Require(payload, at, 8 + gfxRecordSize, "numbered-save range transform");
         long rangeFirst = ReadInt(payload, at);
         int rangeCount = ReadInt(payload, at + 4);
@@ -170,7 +179,25 @@ public static class NativeNumberedSaveCodec
         return new NativeNumberedSaveState(
             ReadInt(payload, 4), ReadInt(payload, 8), soundEffects, resources, surfaces, frames,
             integers, floats, strings, pointers, pointerStrings, localPointerScratch, objects,
-            rangeFirst, rangeCount, rangeRecord);
+            rangeFirst, rangeCount, rangeRecord, legacyTightGfxLayout);
+    }
+
+    private static (NativeSavedGfxObject[] Objects, int After) ReadGfxObjects(
+        ReadOnlySpan<byte> payload, int offset, int count, int strideBytes)
+    {
+        var objects = new NativeSavedGfxObject[count];
+        int at = offset;
+        for (int i = 0; i < objects.Length; i++)
+        {
+            Require(payload, at, 4 + NativeNumberedSaveState.GfxRecordSize,
+                "numbered-save gfx object");
+            long handle = ReadInt(payload, at);
+            objects[i] = new NativeSavedGfxObject(
+                handle,
+                payload.Slice(at + 4, NativeNumberedSaveState.GfxRecordSize).ToArray());
+            at = checked(at + strideBytes);
+        }
+        return (objects, at);
     }
 
     public static NativeNumberedSaveState Empty(IReadOnlyList<NativeSavedScriptFrame> frames)

engine/Age.Engine/Sys4/RetainedSurfaceRasterizer.cs

@@ -0,0 +1,65 @@
+using Age.Engine.Model;
+
+namespace Age.Engine.Sys4;
+
+/// <summary>Platform-neutral software publication of retained gfx objects into an AGE surface.
+/// Native scripts use the same object list for the backbuffer and selected offscreen render targets.</summary>
+public static class RetainedSurfaceRasterizer
+{
+    public static int CompositeRange(
+        RgbaImage destination,
+        IReadOnlyList<RenderObject> visible,
+        long firstHandle,
+        long count,
+        Func<RenderObject, RgbaImage?> resolveSource)
+    {
+        ArgumentNullException.ThrowIfNull(destination);
+        ArgumentNullException.ThrowIfNull(visible);
+        ArgumentNullException.ThrowIfNull(resolveSource);
+        if (destination.Width <= 0 || destination.Height <= 0
+            || destination.Pixels.Length != checked(destination.Width * destination.Height * 4)
+            || count <= 0)
+            return 0;
+
+        int rendered = 0;
+        foreach (RenderObject item in visible)
+        {
+            if (item.Handle < firstHandle || item.Handle - firstHandle >= count) continue;
+
+            TransformState transform = item.Transform;
+            Affine2D localToDestination =
+                Transform2DMath.Build(transform, item.Rotation).FromLocalOrigin(item.DstX, item.DstY);
+            if (item.RangeTransform is { } rangeTransform)
+                localToDestination = localToDestination.Then(rangeTransform);
+
+            RgbaImage? source = resolveSource(item);
+            float opacity = item.Alpha / 255f;
+            float tintStrength = item.TintStrength / 255f;
+            if (source == null)
+            {
+                if (item.SurfaceResId != 0 || item.Blend == BlendKind.Opaque) continue;
+                int width = item.W > 0 ? item.W : 800;
+                int height = item.H > 0 ? item.H : 600;
+                float fillOpacity = item.MultiplyTint ? opacity : opacity * tintStrength;
+                SoftwareAffineRasterizer.FillRgba(
+                    destination.Pixels, destination.Width, destination.Height,
+                    width, height, localToDestination, item.Tint, fillOpacity);
+                rendered++;
+                continue;
+            }
+
+            if (item.W <= 0 || item.H <= 0 || item.SrcX < 0 || item.SrcY < 0) continue;
+            int widthToDraw = System.Math.Min(item.W, source.Width - item.SrcX);
+            int heightToDraw = System.Math.Min(item.H, source.Height - item.SrcY);
+            if (widthToDraw <= 0 || heightToDraw <= 0) continue;
+            SoftwareAffineRasterizer.BlitRgba(
+                destination.Pixels, destination.Width, destination.Height,
+                source.Pixels, source.Width, source.Height,
+                item.SrcX, item.SrcY, widthToDraw, heightToDraw,
+                localToDestination, item.Tint, tintStrength, opacity,
+                item.MultiplyTint, item.Blend);
+            rendered++;
+        }
+        return rendered;
+    }
+}

engine/Age.Engine/Vm/ExecFrame.cs

@@ -1,4 +1,5 @@
 using Age.Engine.Model;
+using Age.Engine.Persistence;
 namespace Age.Engine.Vm;
 
 /// <summary>One script activation: the running script, its instruction cursor, its local slots,
@@ -10,6 +11,10 @@ internal sealed class ExecFrame
 
     public readonly Script Script;
     public int Pc;                                       // entry instruction index
+    // While a restored descendant is running, this activation is parked at the synthetic op-0xae
+    // rendezvous rather than its native T1/T2/T3 coordinate. Retain the serialized coordinates until
+    // the descendant returns so an intervening save can reproduce the native parent frame.
+    public NativeSavedScriptFrame? RestoredSaveFrame;
     public int ReadMessageOffset = -1;                   // latest op-0x71 code DWORD coordinate
     public readonly Frame Locals = new();
     public readonly List<int> CallStack = new();         // intra-script `call` (op 0x8f) returns

engine/Age.Engine/Vm/VirtualMachine.cs

@@ -595,12 +595,17 @@ public sealed class VirtualMachine
         Script root = _s;
         int rootEntry = root.IndexByOffset.TryGetValue(entryOffset, out var idx) ? idx : 0;
         FrameCause cause = FrameCause.TopScene;
+        NativeSavedScriptFrame? restoredRootFrame = null;
         while (true)
         {
             FrameOutcome outcome;
             try
             {
-                outcome = RunFrame(new ExecFrame(root, rootEntry), cause);
+                ExecFrame rootFrame = restoredRootFrame == null
+                    ? new ExecFrame(root, rootEntry)
+                    : CreateRestoredFrame(root, restoredRootFrame);
+                restoredRootFrame = null;
+                outcome = RunFrame(rootFrame, cause);
             }
             catch (NumberedRestoreRequestedException)
             {
@@ -618,8 +623,9 @@ public sealed class VirtualMachine
                     }
                 }
                 root = ResolveSavedScript(_loadedNumberedState.Frames[0]);
-                rootEntry = FindRestoreRendezvous(root);
+                rootEntry = 0;
                 _restoreFrameIndex = 0;
+                restoredRootFrame = _loadedNumberedState.Frames[0];
                 cause = FrameCause.SaveRestore;
                 continue;
             }
@@ -800,19 +806,26 @@ public sealed class VirtualMachine
         for (int i = 0; i <= cutoff; i++)
         {
             ExecFrame frame = active[i];
-            int[] returns = frame.CallStack
+            NativeSavedScriptFrame? restored = frame.RestoredSaveFrame;
-                .Where(returnPc => (uint)returnPc < (uint)frame.Script.Instructions.Count)
+            int[] returns = restored?.ReturnIndices.ToArray()
-                .Select(returnPc =>
+                ?? frame.CallStack
-                {
+                    .Where(returnPc => (uint)returnPc < (uint)frame.Script.Instructions.Count)
-                    int returnOffset = frame.Script.Instructions[returnPc].Offset;
+                    .Select(returnPc =>
-                    return FindTableIndex(frame.Script.LocalCallOffsets, returnOffset - 3);
+                    {
-                })
+                        int returnOffset = frame.Script.Instructions[returnPc].Offset;
-                .Where(index => index >= 0)
+                        return FindTableIndex(frame.Script.LocalCallOffsets, returnOffset - 3);
-                .ToArray();
+                    })
-            int resumeIndex = CurrentReadMessageIndex(frame);
+                    .Where(index => index >= 0)
-            int callTargetIndex = i == cutoff || (uint)frame.Pc >= (uint)frame.Script.Instructions.Count
+                    .ToArray();
+            int resumeIndex = restored?.ResumeIndex ?? CurrentReadMessageIndex(frame);
+            int callTargetIndex = i == cutoff
                 ? -1
-                : FindTableIndex(frame.Script.ScriptCallOffsets, frame.Script.Instructions[frame.Pc].Offset);
+                : restored?.CallTargetIndex
+                    ?? ((uint)frame.Pc >= (uint)frame.Script.Instructions.Count
+                        ? -1
+                        : FindTableIndex(
+                            frame.Script.ScriptCallOffsets,
+                            frame.Script.Instructions[frame.Pc].Offset));
             frames[i] = new NativeSavedScriptFrame(
                 i - 1, frame.Script.PackedId, returns, resumeIndex, callTargetIndex);
         }
@@ -941,6 +954,29 @@ public sealed class VirtualMachine
             $"Saved script {script.Name} has no opcode 0xae restore rendezvous.");
     }
 
+    private static ExecFrame CreateRestoredFrame(Script script, NativeSavedScriptFrame saved)
+    {
+        // Native creates each saved script context at its ordinary entrypoint. The script runs its
+        // local-array/constants/resource prologue and rendezvouses at 0xae itself; jumping directly
+        // to 0xae leaves those frame-local tables zeroed (FIELD then collapses its map zoom to 0%).
+        _ = FindRestoreRendezvous(script);
+        int entry = script.IndexByOffset.TryGetValue(0, out int index) ? index : 0;
+        var frame = new ExecFrame(script, entry)
+        {
+            RestoredSaveFrame = saved,
+        };
+        if ((uint)saved.ResumeIndex < (uint)script.ReadMessageOffsets.Count)
+            frame.ReadMessageOffset = script.ReadMessageOffsets[saved.ResumeIndex];
+        foreach (int returnIndex in saved.ReturnIndices)
+        {
+            if ((uint)returnIndex >= (uint)script.LocalCallOffsets.Count) continue;
+            int returnOffset = checked(script.LocalCallOffsets[returnIndex] + 3);
+            if (script.IndexByOffset.TryGetValue(returnOffset, out int returnPc))
+                frame.CallStack.Add(returnPc);
+        }
+        return frame;
+    }
+
     private static int ResolveTableOffset(
         Script script, IReadOnlyList<int> table, int index, int fallback)
     {
@@ -1039,6 +1075,9 @@ public sealed class VirtualMachine
         {
             case "script-entry":
                 Gfx.ClearSurfaceReloadPolicies(); return pc + 1;
+            case "set-surface-persistence-flags": // 0x258 (slot)(flags): bit 0 = numbered-load reload
+                Gfx.SetSurfaceReloadOnRestore(unchecked((int)Read(a[0])), (Read(a[1]) & 1) != 0);
+                return pc + 1;
             case "add": Write(a[0], Read(a[1]) + Read(a[2])); return pc + 1;
             case "sub": Write(a[0], Read(a[1]) - Read(a[2])); return pc + 1;
             case "mul": Write(a[0], Read(a[1]) * Read(a[2])); return pc + 1;
@@ -1139,6 +1178,11 @@ public sealed class VirtualMachine
                 bool terminal = _restoreFrameIndex == _loadedNumberedState.Frames.Count - 1;
                 if (terminal)
                 {
+                    // Native restores the saved top frame as the numbered-save boundary selected by
+                    // opcode 0x1ad. Re-establish that identity so a subsequent save excludes transient
+                    // SAVE/menu helper frames instead of serializing the currently open modal stack.
+                    lock (_debugControlLock) _saveResumeFrame = _cur;
+                    _cur.RestoredSaveFrame = null;
                     _loadedNumberedState = null;
                     _restoreFrameIndex = -1;
                     return ResolveTableOffset(_cur.Script, _cur.Script.ReadMessageOffsets, saved.ResumeIndex, pc + 1);
@@ -1149,11 +1193,14 @@ public sealed class VirtualMachine
                 Script child = ResolveSavedScript(childSaved);
                 _restoreFrameIndex = parentIndex + 1;
                 FrameOutcome childOutcome = RunFrame(
-                    new ExecFrame(child, FindRestoreRendezvous(child)), FrameCause.SaveRestore,
+                    CreateRestoredFrame(child, childSaved), FrameCause.SaveRestore,
                     childSaved.ScriptId);
                 _restoreFrameIndex = parentIndex;
-                if (childOutcome is FrameOutcome.Halted or FrameOutcome.ExitRequested)
+                if (childOutcome == FrameOutcome.Halted) return HALT;
-                    return HALT;
+                if (childOutcome == FrameOutcome.RootReload) return ROOT_RELOAD;
+                if (childOutcome == FrameOutcome.ExitRequested)
+                    throw new ProcessExitRequestedException();
+                _cur.RestoredSaveFrame = null;
                 return ResolveTableOffset(
                     _cur.Script, _cur.Script.ScriptCallOffsets, saved.CallTargetIndex, pc) + 1;
             }

godot/GodotAdvHost.cs

@@ -274,6 +274,11 @@ public sealed class GodotAdvHost : IHost
 
     public void PresentObjectRange(GfxState gfx, long firstHandle, long count)
     {
+        if (gfx.CurrentRenderTargetSlot >= 0)
+        {
+            PublishObjectRangeToSurface(gfx, firstHandle, count);
+            return;
+        }
         System.Threading.Interlocked.Exchange(ref _presentRequested, 1);
         _timeline?.Event("present-object-range", new() { ["first"] = firstHandle, ["count"] = count });
     }
@@ -573,6 +578,7 @@ public sealed class GodotAdvHost : IHost
         {
             int slot = gfx.CurrentRenderTargetSlot;
             var snapshot = gfx.SnapshotVisibleObjects(_clock.NowMs);
+            PublishObjectRangeToSurface(gfx, 0, long.MaxValue, snapshot);
             lock (_screenTransitionLock) _renderTargetSnapshots[slot] = snapshot;
             _timeline?.Event("render-target-snapshot", new()
             {
@@ -1110,10 +1116,67 @@ public sealed class GodotAdvHost : IHost
         // For an offscreen target, discard separately retained text draws so its modeled pixel contents
         // observe the native D3D clear as well.
         if (surfaceSlot >= 0)
+        {
             lock (_textLock) _surfaceText.Remove(surfaceSlot);
+            lock (_imageLock)
+            {
+                if (_surfaceImages.TryGetValue(surfaceSlot, out var image))
+                    System.Array.Clear(image.Pixels);
+                else if (_slotDims.TryGetValue(surfaceSlot, out var dimensions)
+                         && dimensions.W > 0 && dimensions.H > 0)
+                    _surfaceImages[surfaceSlot] = new RgbaImage(
+                        dimensions.W, dimensions.H,
+                        new byte[checked(dimensions.W * dimensions.H * 4)]);
+            }
+        }
         _timeline?.Event("render-target-clear", new() { ["surface"] = surfaceSlot });
     }
 
+    private void PublishObjectRangeToSurface(
+        GfxState gfx, long firstHandle, long count,
+        IReadOnlyList<RenderObject>? sampled = null)
+    {
+        int targetSlot = gfx.CurrentRenderTargetSlot;
+        if (targetSlot < 0 || !_slotDims.TryGetValue(targetSlot, out var dimensions)
+            || dimensions.W <= 0 || dimensions.H <= 0)
+            return;
+
+        RgbaImage destination;
+        lock (_imageLock)
+            destination = _surfaceImages.TryGetValue(targetSlot, out var current)
+                ? new RgbaImage(current.Width, current.Height, (byte[])current.Pixels.Clone())
+                : new RgbaImage(dimensions.W, dimensions.H,
+                    new byte[checked(dimensions.W * dimensions.H * 4)]);
+
+        IReadOnlyList<RenderObject> visible = sampled ?? gfx.SnapshotVisibleObjects(_clock.NowMs);
+        int rendered = RetainedSurfaceRasterizer.CompositeRange(
+            destination, visible, firstHandle, count,
+            item =>
+            {
+                var raw = gfx.TryGet(item.Handle);
+                var resolved = raw != null
+                    ? ResolveSurfaceTexture(raw.SourceSlot, item.SurfaceResId)
+                    : ResolveResIdTexture(item.SurfaceResId);
+                return resolved == null
+                    ? null
+                    : RgbaSurfaceOps.WithColorKey(resolved.Value.Image, item.ColorKey);
+            });
+
+        lock (_imageLock) _surfaceImages[targetSlot] = destination;
+        IReadOnlyList<RenderObject> retained = visible
+            .Where(item => item.Handle >= firstHandle && item.Handle - firstHandle < count)
+            .ToArray();
+        lock (_screenTransitionLock) _renderTargetSnapshots[targetSlot] = retained;
+        _timeline?.Event("render-target-publish", new()
+        {
+            ["surface"] = targetSlot,
+            ["first"] = firstHandle,
+            ["count"] = count,
+            ["objects"] = retained.Count,
+            ["rendered"] = rendered,
+        });
+    }
+
     public void ReleaseSurfaceRange(int firstSlot, int count)
     {
         IReadOnlyList<MovieSurfaceBinding> stoppedMovies = _movieSurfaces.ReleaseRange(firstSlot, count);

tools/age_opcodes_himegari.py

@@ -76,6 +76,5 @@ INFERRED: dict[int, dict] = {
     0x239: dict(name='animate-gfx-srcrect-target', category='draw', noop=False, confidence='high', source='investigation', summary='(handle)(delay_ms)(duration_ms)(frame_count)(column_count)(target_frame) — one-shot row-major source-rectangle cell channel. Worker gfx_worker_set_srcrect_cell @0x47ed90 stores timing at obj+0x48/+0x5c, layout at +0x238/+0x23c, and target at +0x234. C# currently retains the endpoint cell immediately.'),
     0x23b: dict(name='draw-decimal-glyphs', category='draw', noop=False, confidence='high', source='investigation', summary='Draw an integer as decimal glyph objects from a style registered by opcode 0x13a.'),
     0x23f: dict(name='query-surface-stop-time-ms', category='draw', noop=False, confidence='high', source='investigation', summary='(out_stop_time_ms)(surface_slot) — query the DirectShow stop position retained by a loaded movie surface, convert seconds to integer milliseconds by truncating toward zero, and write -1 when the movie slot is empty. Port-only host decoder failure is modeled as an explicitly completed, zero-duration movie.'),
-    0x258: dict(name='decl?', category='marker', noop=True, confidence='low', source='harness', summary='2 imm; runs in a chain right after script-entry 0x259, enumerating ids — prologue declaration/registration?'),
     0x2c5: dict(name='byte-string-length', category='compute', noop=False, confidence='high', source='investigation', summary="Write the resolved NUL-terminated engine string's raw byte length."),
 }

vm-map/opcodes.toml

@@ -1724,12 +1724,12 @@ abi_source = "kelebek+decode-validated"
 name = "continue-save-load-stack-restore"
 category = "control"
 summary = "() - during serialized save restoration, replace the current frame PC with its saved resume/call target and advance through the saved script-context stack; otherwise a no-op."
-details = "Layout 3 frame d259 indexes SYS4 T1 read-message reset sites, d260 indexes T2 call-script sites, and the saved local return stack indexes T3 local-call sites. Port status (2026-07-24): the active path reconstructs the saved recursive frame chain and resumes the terminal frame at its T1 boundary. The installed SAVE00 continuation gate proves SYSTEM4 -> FORT restoration reaches FORT's CHMENU gameplay poll; the synthetic gate asserts the child frame enters with SaveRestore rather than ordinary CallScript cause."
+details = "Layout 3 frame d259 indexes SYS4 T1 read-message reset sites, d260 indexes T2 call-script sites, and the saved local return stack indexes T3 local-call sites. Port status (2026-07-24): each saved script is loaded at its ordinary entry, runs its frame-local prologue, and reaches 0xae itself; the active path then reconstructs the saved recursive frame chain, preserves each synthetic ancestor's original T1/T2/T3 coordinates while its restored child is active, reinstates the terminal restored frame as opcode 0x1ad's save boundary, and resumes it at T1. The installed SAVE00 continuation gate proves SYSTEM4 -> FORT restoration reaches FORT's CHMENU gameplay poll; the unchanged port-authored dungeon slot 006 proves FIELD's pre-rendezvous zoom table is initialized; synthetic gates assert both root/child prologue execution, SaveRestore rather than ordinary CallScript entry, exclusion of nested SAVE helpers from a re-save, and reload of that rewrite without ordinary boot re-entry."
 noop_headless = false
 source = "investigation"
 confidence = "high"
 depends_on = []
-evidence = "Ghidra /v2: op_0xae_continue_save_load_stack_restore@0x416790 first tests ctx+0x53d24 (set by save_data_deserialize_and_begin_restore@0x40fd10). When clear it returns. When set, it selects the serialized frame layout through set:SaveVersion1/2, restores the current PC from that layout's saved return/call target, advances through contexts with script_frame_restore_saved_layout@0x40f2d0, and clears the restore flag on reaching the saved terminal context. Its 305 corpus sites overwhelmingly follow coroutine-resume/call boundaries, which provide the rendezvous points used while reconstructing the stack."
+evidence = "Ghidra /v2: op_0xae_continue_save_load_stack_restore@0x416790 first tests ctx+0x53d24 (set by save_data_deserialize_and_begin_restore@0x40fd10). When clear it returns. When set, it selects the serialized frame layout through set:SaveVersion1/2, restores the current PC from that layout's saved return/call target, advances through contexts with script_frame_restore_saved_layout@0x40f2d0, and clears the restore flag on reaching the saved terminal context. The restore helper calls script_frame_load_resource@0x40e980, which initializes each new frame PC to its script codebase; the script therefore executes its ordinary prologue before reaching 0xae. Native parent contexts retain restored coordinates while the child runs. SAVE00 -> SAVE03 proved the port must retain the terminal 0x1ad boundary; SAVE00 -> SAVE04 proved a managed parent parked at synthetic 0xae must serialize original resume=8/call=8 rather than synthetic -1/-1; slot 006 proved direct entry at 0xae skips FIELD's zoom-table prologue and collapses the dungeon map. Its 305 corpus sites overwhelmingly follow coroutine-resume/call boundaries."
 
 [[opcode]]
 op = 0xb4
@@ -5742,12 +5742,12 @@ abi_source = "kelebek+decode-validated"
 [opcode.semantics]
 name = "present-gfx-object-range"
 category = "draw"
-summary = "(first_handle)(count) - flush/present retained graphics objects in the selected handle range and clear their pending update flags."
+summary = "(first_handle)(count) - flush/present retained graphics objects in the selected handle range into the currently selected backbuffer or offscreen render target, then clear their pending update flags."
 noop_headless = false
 source = "investigation"
 confidence = "high"
 depends_on = []
-evidence = "Ghidra /v2: op_0x222_handler@0x4235e0 calls gfx_present_object_range@0x482230. The worker enters the graphics service, walks the retained-object map, processes flagged objects whose handles fall in [first,first+count), clears pending flags, and finalizes the render batch. HISTORY.BIN uses (0,60000) after rebuilding its retained presentation."
+evidence = "Ghidra /v2: op_0x222_handler@0x4235e0 calls gfx_present_object_range@0x482230. The worker enters the graphics service, walks the retained-object map, processes flagged objects whose handles fall in [first,first+count), clears pending flags, and finalizes the render batch in the D3D target previously selected by op 0x20d. HISTORY.BIN uses (0,60000) for the backbuffer. SAVE.BIN instead renders [0,0x130b0) into 800x600 slot 2, then handle 0x15f90 at 14% scale into 112x84 slot 192; op 0x1ae writes slot 192 as the numbered .STH thumbnail."
 
 [[opcode.semantics.args]]
 i = 1
@@ -6709,28 +6709,29 @@ observed_types = ["imm"]
 
 [[opcode]]
 op = 0x258
-label = "u00422FE0"
+label = "set-surface-persistence-flags"
 argc = 2
 abi_source = "kelebek+decode-validated"
 
 [opcode.semantics]
-name = "decl?"
+name = "set-surface-persistence-flags"
-category = "marker"
+category = "draw"
-summary = "2 imm; runs in a chain right after script-entry 0x259, enumerating ids — prologue declaration/registration?"
+summary = "(surface_slot)(flags) - replace the native surface record's numbered-save persistence flags; bit 0 controls asset reload on restore and bit 1 controls the adjacent still-unnamed field."
-noop_headless = true
+details = "Scripts place declaration chains immediately after opcode 0x259 clears both fields. The port models bit 0 because it is consumed by layout-3 restoration; bit 1 is retained as an identified native field but has no known runtime consumer yet."
-source = "harness"
+noop_headless = false
-confidence = "low"
+source = "investigation"
+confidence = "high"
 depends_on = []
-evidence = ""
+evidence = "Ghidra /v2: op_0x258_set_surface_persistence_flags@0x4250a0 reads the slot and flags operands and calls gfx_surface_set_persistence_flags@0x4159c0. The worker writes flags&1 to record +0x08 and (flags>>1)&1 to +0x0c in both 1,000-record tables. This corrects the old upstream address/name association: 0x422fe0 is opcode 0x20f movie playback, not 0x258. Corpus declaration chains follow opcode 0x259 at script entry."
 
 [[opcode.semantics.args]]
 i = 1
-role = ""
+role = "surface slot"
 observed_types = ["imm"]
 
 [[opcode.semantics.args]]
 i = 2
-role = ""
+role = "persistence flags"
 observed_types = ["imm"]
 
 [[opcode]]