- README: attract-mode stills + demo GIF as the hero row, status/warning
callouts, metrics table, "Contents" index, doc-index table; simplified
the two-agent workflow diagram (LR -> TD, fewer edges) for GitHub web
- Table-of-contents indexes added to internals.md, porting-notes.md,
agentic-development.md, AUDIO-PLAN.md
- CI: add Windows/MSYS2 build job, minimal token (contents: read),
persist-credentials: false, concurrency group; run on main + master.
validate job regenerates the US ROM-asset symbol file only and configures
ntsc-final (PAL/JP need a region ROM; no ROM is used in CI)
- .github: SECURITY.md, issue forms + chooser config, PR template,
labels.yml, dependabot (Actions, monthly)
- Do not track the proprietary SGI IRIX IDO toolchain or SGIImageViewer.exe;
document in SetupGuide.md / building.md how to obtain IDO for N64 ROM
builds (the PC port does not need it)
- Stop tracking two committed *.pyc files; drop stale screenshots and a
redundant .gitignore line; misc doc cross-reference fixes
Claude-Session: https://claude.ai/code/session_01GULf6knbsqfsnW39tMr9WT
Pre-spawn intro-camera views captured with GE_PCDUMP after the D168
flip fix (so they're upright). Not yet wired into the README - awaiting
placement direction (keep archives-1 / bunker1-1 as the top pair).
- attract-bunker1.png: control room + "Satellite Control Bunker" caption
- attract-silo.png: guards + toxic-waste drums, silo hangar behind
- attract-dam.png: guards on patrol, control panel, snowy peaks
Claude-Session: https://claude.ai/code/session_01UpTM1AcBRJuTetoZNcCD5V
Built the migrated tree (MSYS2 mingw64 + ROM + sidecar regen) and ran
-level_09 with GE_PCDUMP: HUD text and the exit arrow now read
right-side-up. D114/D116 confirmed as capture artifacts.
- tools_pc/golden/*.png regenerated from a fresh correctly-oriented run
(replaces the stopgap in-place flip); stray v*.png removed
- D75 shrunk: with upright captures the gun-barrel Bond RENDERS fine
(walk + fire) -- the "absent" reports were the flipped capture. The
Nintendo logo is genuinely broken (renders as 2 white blobs shifted
left -- a real logoinst transform bug, not a flip). Cast roll not
re-captured.
- D169 entry -> FIXED with the build/verify note
Claude-Session: https://claude.ai/code/session_01UpTM1AcBRJuTetoZNcCD5V
- Commit CLAUDE.md + AGENTS.md so a fresh clone has the operating manual and
the ABI/no-game-logic rules (trimmed of stale phase status; the personal
pd_port path removed; points at the split docs).
- Sweep every stale doc reference across the tree after the restructure:
PCPortResearch.md -> internals.md / dev/findings.md,
PORT-LEARNINGS.md -> porting-notes.md,
AGENT-WORKFLOW.md -> dev-process.md,
docs/{GRAPHICS-BACKLOG,HANDOFF-ARCHIVE,GE-ENV-PROBES,LEVEL-*,TEXTURE-GLITCH-ANALYSIS,AUDIT-M6-*,AUDIO-PLAN}.md -> docs/dev/.
- Drop the superseded docs/PLAN-linear-level-sweep.md.
- docs/HANDOFF.md + docs/BACKLOG.md stay local-only (rolling working files);
CLAUDE.md now handles their absence in a fresh clone.
Front-end mouse pointer can't reach the outer mission-select grid cells
(only an inner ~3x3). The D165 P-controller in port/src/input.c hard-codes
a 320x240 virtual field; GE's front end runs 440x330, so the pointer
target/estimate saturate at 300/220 before the grid's outer split points
(317 / 235.5). File-select/mode/main-menu unaffected (targets inside the
320x240 sub-box). Fix proposed (derive clamp bounds from
getPlayer_c_screenwidth/height, seed estimator from the real cursor) but
NOT applied - no build in the migrated tree to verify link + regression.
Claude-Session: https://claude.ai/code/session_01UpTM1AcBRJuTetoZNcCD5V
gfx_opengl_dump_bound_fbo wrote glReadPixels output (GL bottom-left
origin) straight into a top-row-first P6 PPM, so every GE_PCDUMP / F12
capture was upside-down. This is the entire source of the D114/D116
"HUD/text X-mirror" - four sessions chasing a fast3d mirror that never
existed, misreading inverted screenshots of asymmetric content. The
developer confirms the game renders correctly on real hardware.
- writer now emits rows top-to-bottom
- tools_pc/golden/*.png flipped in place to match + golden/README.md
- D114/D116 reclassified as capture-orientation artifacts (not deleted);
D75 "logo mispositioned" / RC1 "photo 180 deg" flagged for re-judging
- annotations in findings.md, GRAPHICS-BACKLOG.md,
TEXTURE-GLITCH-ANALYSIS.md, porting-notes.md D2, GE-ENV-PROBES.md
Not runtime-verified (migrated tree has no ROM/toolchain) - needs a
fresh capture to confirm text reads normally.
Claude-Session: https://claude.ai/code/session_01UpTM1AcBRJuTetoZNcCD5V
- README: "not playable yet" callout under the screenshots (no audio,
broken front-end models/text, expect crashes past the level intro).
- README/building.md/ci.yml: platform testing is Windows-only. The Linux
build is provided but has had zero functional testing at runtime; the
CI job is a CMake-configure smoke test, not a Linux-support claim.
Claude-Session: https://claude.ai/code/session_01UpTM1AcBRJuTetoZNcCD5V
- README, LICENSE (MIT for the port layer), NOTICE, CONTRIBUTING.
- Restructure docs: internals.md, porting-notes.md, dev-process.md +
dev/agentic-development.md (the research write-up), dev/ (raw record).
- Replace decomp status-badge CI with a PC-port validate workflow.
- Remove session/agent scaffolding and stale planning docs from the tree.
- Add .gitattributes (normalize line endings to LF); add screenshots.
The pre-2026-08 history is the GoldenEye 007 decompilation by Larry Ficken
(KholdFuzion) and contributors, preserved unchanged. The PC-port work from
2026-08-17 on is mine.
Adds env-gated GE_D75 probe in title.c sub_GAME_7F007F30 (capped 8 lines).
Booted the bare front end through the gun-barrel sequence:
- Bond chr model + gun model entirely absent; both Model instances are
valid (non-NULL obj, numMatrices 21/1) -> 'model not loaded' ruled out.
- chrModelInstance->render_pos is valid and freshly written each frame
(== that frame's dynAllocate mtxlist base, clean double-buffer swap,
o2p returns it unchanged) -> the 'render_pos = stale dynAllocate arena'
hypothesis (D115 #5) is RULED OUT; do NOT land a persistent render_pos
buffer.
- Zero fast3d DL warnings during the sequence -> not the D144/D146
corrupt-sub-DL family.
Failure is downstream in drawjointlist/dotube vtx/node-DL resolution or an
off-screen basemtx. Next step: a drawjointlist-level probe. Full write-up
in PCPortResearch.md 'D75 Bug 2 - RUNTIME PROBE'.
-level_09 nonclear stable (91.66->91.62, D117 pan noise); build green.
Static-only pass. (a) D73 scope gap DISPROVEN: src/libultra/gu/* and
matrixmath.c are all endian-clean; the Rareware logo exercises the full
gu pipeline front-end and renders correctly.
D75 now splits into two independent bugs:
- Bug 1 (Nintendo/GE logo transform, wallet/Bond-photo 180deg): the
parked D114/D116 shared fast3d viewport/MP-matrix mirror. Not game
code; static tracing banned per PORT-LEARNINGS D2.
- Bug 2 (gun-barrel Bond + cast models entirely absent): category (b),
independent of any transform. In-level skeletal models render via the
identical joint path, so the discriminator is the front-end setup:
render_pos pointed at a transient dynAllocate gfx arena (D115 MED #5)
is the lead suspect and an in-scope #ifdef PORT fix candidate.
docs only: PCPortResearch.md D75 ADDENDUM + index row; PORT-LEARNINGS
section C (D73/gu scope now settled - stop re-auditing).
fast3d never stored the N64 tile mask and wrapped textures purely at the
uploaded image size; the RDP wraps a WRAP render tile at 1<<mask, which GE
sets to ceil(log2(dim)) - so non-power-of-two surfaces (Depot room
textures) repeat at the next power of two on console and one image-width
too soon in the port ('textures repeat oddly').
- gfx_dp_set_tile: store masks/maskt on the tile.
- gfx_sp_tri1 hoisted pre-wrap block (the reworked D74 block, indexed by
texunit): when WRAP and 1<<mask != tex_width, fold UV at 1<<mask and
clamp the [dim,1<<mask) overflow band (no real texels) to the edge.
- GE_WRAPFIX env override (wins over the ini) for testing.
Behind the existing Video.WrapFix knob, default OFF -> byte-identical to
golden. Per-level captures (-level_09/-30/-34/-20, off vs on): no crashes,
no structural regression; Depot shows small localized ceiling/wall texel
changes. Kept default off pending a human visual check on Depot.
docs: PCPortResearch.md RC3/D167, PORT-LEARNINGS.md D, TEXTURE-GLITCH-ANALYSIS.md 6d.
Static audit of every osSetIntMask(OS_IM_NONE) in the compiled audio code
(event.c, csplayer.c, synaddplayer.c, snd.c): all balanced on every exit
path, including alEvtqPostEvent's freeList-empty early return. No unbalanced
same-thread leak exists -- the D152 s_imLock wedge is not a missing restore.
Two real defects fixed (#ifdef PORT, N64 verbatim under #else):
1. snd.c sndSetSfxSlotVolume walked the live ALSoundState list and posted to
the shared ALEventQueue without holding OS_IM_NONE, unlike its twin
sndDeactivateAllSfxByFlag. On PC: unguarded walk racing amMain + one lock
acquire/release per matching sound. During the mission-failed fade-out
(sndSetScalerApplyVolumeAllSfxSlot -> sndApplyVolumeAllSfxSlot -> here,
per frame) that is a per-frame lock-acquire storm. Now holds the mask once
across the walk; sndApplyVolumeAllSfxSlot batches its slot loop under one
recursive hold.
2. port/src/libultra.c: portThreadWrapper now calls imThreadExitRelease() when
a host thread returns from its entry -- if it still owns s_imHeld, release
the orphan (+ LOG_ERROR). Kills the 'transient thread acquired OS_IM_NONE
and exited' leak (the §F dump's own alternate hypothesis) and the re-wedge
when a new thread reuses the dead pthread id and imAcquire mis-detects
recursion. Steal-lock kept as backstop.
Build 240/240 green; -level_09 frame 300 renders 91.7% non-clear crash-free;
-level_20 boots crash-free. Fade-out repro is playtest-gated (not headless-
reproducible) -- user to confirm no black screen + no D152: log lines.
Re-audited the M-28 bgTestRayIntersectionInRoom GBI-parser port (bg.c):
- G_TRI1 (w1>>16/8/0 & 0xff /10) and all 12 G_TRI4 nibble recoveries
cross-checked twice against the N64 #else path and the raw BE-word bit
layout -- both CORRECT, no change.
- BUG: vtxoff read gdl->dma.par & 0xf. The PC Gdma_le shim maps .par to
bits 0-23 of word0 (packed length, 16-aligned) not the N64 byte-1
params field at bits 16-23, so vtxoff was silently 0 -> wrong
vtxbase[idx] for any G_VTX batch with v0 != 0. Now
((u32)gdl->words.w0 >> 16) & 0xf.
- Sibling ported: bgTestBulletHitBackground tail (post-hit G_SETTILE
back-scan for tileformat/tilesize) was itself an unported raw
((u8*)gdl)[0]/[1] + 64-bit words.w0<<11>>30 parser over the same
D85-widened 16-byte room DL. #ifdef PORT: opcode from bits 24-31,
byte1 from bits 16-23, (u32)-forced 32-bit width for the tilesize
shift. N64 path kept under #else.
- Diagnostic: GE_D154=1 logs the parsed cmd / vtx-index stream (call
header + per-TRI1/TRI4 idx) for the first 64 invocations so the
integrator/user can eyeball the parse. Verification is playtest-gated
(shoot a wall/floor on an idle machine) -- the no-input level sweep
never enters this function.
Build green (ntsc-final, 242/242). ABI/GBI-parse only, no game-logic
change. docs/PCPortResearch.md D154 + PORT-LEARNINGS §B updated.
port/src/input.c only (port-layer, no game-code change):
D165 - the M-30 menu 'pointer mode' fed mouse velocity into the stick, but
front.c frontUpdateControlStickPosition integrates the stick as a velocity
into a screen-pixel cursor -> velocity^2 feel. Replace with a P-controller:
maintain menuEst{H,V} (our estimate of the game cursor, stepped each poll with
front.c's exact recurrence via menuCursorStep), accumulate menuTgt{H,V} from
raw mouse motion, emit stick = clamp(6.0*(target-est), +/-70). Gain 6.0 =>
per-poll error decay 0.45, no overshoot, ~8-poll settle, parks under the
game deadzone with no drift. Input.MenuPointerMode (default 1) keeps the old
velocity mode as a fallback; Input.MenuPointerSpeed scales the mapping;
GE_INPUTLOG prints an est/tgt/eff/stick trace. No mouse motion => emitted
stick unchanged => prior behaviour exactly.
D166 - hipfire mouse pitch was a fixed digital threshold while yaw is analog.
Emit the C-button as a duty-cycled pulse whose frequency scales with mouse-Y
speed (fast = solid hold as before, slow = sparse taps). Input.HipfirePitchSpeed
tunes it. Aim-mode (RMB, already analog) untouched.
Build green; -level_09 (~360 frames) + GE_STARTMENU=6 smoke crash-free.
Feel-check owed (headless can't drive the mouse).
The Depot roof (CI8 tile drawn with gsDPSetTextureLUT(G_TT_NONE)) was
decoded against a stale rdp.palette, producing blue/magenta noise that
aliased into radial rays on the receding ceiling plane. A CI-siz tile
with the TLUT disabled must be treated as raw intensity (I). Route
CI4/CI8 -> I4/I8 in import_texture() when palette_fmt == G_TT_NONE.
Adds GE_TEXDUMP (env-gated PPM + param dump) as an inert diagnostic.
-level_09/-30/-34 sweep PASS, unregressed.
D159 (RC1 / D149): texSwapAltRowBytes pre-swaps odd texture rows in u32
groups to cancel the N64 RDP odd-line TMEM address XOR during word block
loads. fast3d emulates no such XOR, so odd rows uploaded scrambled -> the
'interlaced' wallet-Bond photo and other large front-end I4/I8/IA8
textures. #ifdef PORT early-return; N64 body kept verbatim. Wallet photo
now renders as a clean grayscale portrait (still 180deg-rotated - D75
model-transform family, parked). -level_09 frame 200 coverage unregressed
(91.62 vs 91.66 golden = D117 noise), visual frame clean.
D164 (D76): constructor_menu00_legalscreen bounded its text loop on
&legalscreen_MRD assuming linker adjacency to legalpage_text_array;
mingw reorders file-scope globals (MRD lands 0x60 before the array) so
the do/while ran once. #ifdef PORT bounds on the array's own length.
d43_emit.py reserved the PointUsage region (add_region) during layout but
the emit pass never wrote it -> all zeros -> chr.c's blood-decal walk
'index = PointUsage[index]' cycled 0->0 forever (capped by the #ifdef
PORT guard at chr.c:3328, so decals were missing/wrong, not a hang).
Now byteswaps each s16 index (numVertices entries, negative-terminated,
no remap). Full regen chain re-run, round-trip validation passes. The
chr.c guard stays as a safety net (SMALL-FIXES B3). Also: retract RC4
(palette off-by-one) in TEXTURE-GLITCH-ANALYSIS -- the current RGBA5551
decode is correct; the doc's 'spec' was ARGB1555.
Reported: menu / room textures look 'interlaced' (alternating garbage
rows). Root cause (docs/TEXTURE-GLITCH-ANALYSIS.md RC2): GE's
texGetDepthAndSize() packs base level + every LOD mip into one
gDPLoadBlock, so fast3d's upload height (size_bytes / row) runs ~1.3x
taller than the base image and the mip bytes render as rows below it.
GE_DTEX=1 confirmed it: 64x64 I4 uploaded 64x87, 65x65 -> 65x95, etc.
Fix in import_texture(): when LOD is active (rdp.tex_lod) and the load
is a plain full-width block (not a windowed gfx_dp_load_tile - the D74
sub-tile / Rare-logo case, which is lod=0), clip size_bytes to the
SETTILESIZE base-tile height. GL then builds correct mips from a correct
base. New knob Video.FixMipTextures (default 1; =0 restores old upload,
verified byte-identical to golden -level_09 3/3). BUNKER1 with the fix
on: side-wall textures visibly cleaner, geometry/HUD unchanged; Silo
unregressed. -level_09/-level_20 framediff WILL flag with the fix on
(textures changed for the better) - refresh golden after the user
confirms the menu + Depot look right.
New GE_DTEX=1 env probe logs load params for the first 64 textures.
After D155 the user re-ran the playthrough and it hung again at the end
of Facility -- same stack but frames=12962 frozen CONSTANT for minutes =
a true infinite loop, not D155's slow spiral.
The `while (1)` at model.c:3131 in modelSetAnimFrame2WithChrStuff steps
root-motion accumulation one anim frame at a time from framea to frameb.
frameb comes from modelTickAnim's `frame += playspeed * speed` per tick.
A cutscene anim transition with a near-zero blend/timespeed/unkb0
denominator (model.c:3436/:3477, guarded >0 but not against tiny values)
makes model->speed/playspeed huge or NaN -> frameb huge/NaN ->
floorFloatToInt garbage -> endframe-curframe ~= 2^31 -> effectively
infinite.
#ifdef PORT guards (N64 unchanged):
- modelSetAnimFrame2WithChrStuff: if frameb non-finite or |frameb| >= 1e6,
snap to framea (loop does ~0 iterations; modelConstrainOrWrapAnimFrame
at :3211 still clamps model->frameb to a real frame).
- modelTickAnim: after the tick loop, if frame/frame2 went non-finite,
fall back to the pre-loop model->animframe1/animframe2 so the NaN can't
poison model state every frame.
`!(x > -1e6f && x < 1e6f)` catches NaN too. Anim frames are 0..few-hundred
so 1e6 is a safe ceiling.
Likely deeper cause (needs the repro): these speed/blend fields are Model
members -- a D100/D140-style struct-pun misalignment would feed garbage
exactly when a cutscene sets an unusual transition. Dump
model->speed/playspeed/timespeed/unkb0 in modelTickAnim during the
cutscene if it still hangs.
Verified: build green; -level_09 frame 200 unregressed (guards never fire
in normal play). PORT-LEARNINGS §E.
User hit a "hang" (process alive, heartbeat dumps) at the end of the
Facility outro cutscene. Symbolicated mainThread stack: parked in
modelConstrainOrWrapAnimFrame <- modelTickAnim <- chrUpdateAnim <-
chrTick <- playerTick <- propsTick <- lvlRender across every dump.
Root cause: waitForNextFrame() (src/game/frametiming.c) passes
nextFrameTime = (osGetCount()-prev+387937)/775875 to updateFrameCounters
UNCLAMPED. osGetCount() is wall-clock on the port (D117), not VI-locked,
so a real-time stall (asset load at the cutscene->debrief boundary, made
worse by a local LLM hammering the box) makes nextFrameTime balloon to
hundreds/thousands. That becomes speedgraphframes -> g_ClockTimer, which
drives modelTickAnim's `while (numticks-- > 0)` per character per render
plus dozens of `for (i=0;i<g_ClockTimer;i++)` sim loops -> one frame does
thousands of anim ticks x N chars -> multi-second frame -> heartbeat
trips, and the slow frame feeds an even bigger delta -> unrecoverable
spiral. The N64 was VI-bound and never produced deltaFrames > ~2.
Fix: #ifdef PORT clamp nextFrameTime to FRAMETIMING_PORT_MAX_CATCHUP (6)
in waitForNextFrame(). N64 path verbatim. Timing-compensation class,
cf. D117/D134. PORT-LEARNINGS §E.
Verified: build green (ntsc-final); -level_09 unregressed (frame 200
nonclear delta 0.04pp, full-window crash-free). The clamp only engages
when a frame took >~100ms, so normal play is byte-identical.
The sm64ex 3-point filter shader path already existed in gfx_opengl.cpp but
was never reachable from config, and its min-filter row was all GL_NEAREST
(no mip sampling) so it would shimmer badly.
- gfx_opengl.cpp: 3-point min-filter row -> GL_LINEAR_MIPMAP_{NEAREST,LINEAR}
so minification uses the mips it already force-generates at upload;
anisotropy default 0 -> 4 (0 was invalid for GL_TEXTURE_MAX_ANISOTROPY).
- video.c: Video.TextureFilter now accepts 2 = 3-point + trilinear.
DEFAULT STAYS 1 (bilinear) -- 3-point is more console-authentic but
softens textures at normal viewing distance, and (tested) did NOT fix
the Depot roof garbage. That is a separate texture-data/UV/effect bug,
not filtering -- BRIEF-B2 updated, next step is a GE_DTEX probe.
Verified: default (bilinear) -level_09 frame 320 unregressed vs golden
(nonclear delta 0.00pp, phash 11); build green; 1200-frame -level_30 run
crash-free with 3-point active.
HANDOFF: M-29 session summary + the 3-track plan (A=playtest gate,
B=audio DEFERRED, C=solo quick wins in priority order).
BRIEF-B2-depot-textures.md: captured evidence + leading hypothesis that
the Depot roof garbage is mostly BACKLOG B1 (missing filtering/mips),
not a decode bug; decode path audited clean.
bgTestRayIntersectionInRoom (bg.c) is the D135 sibling flagged since M-23:
an unported N64 GBI parser walking the 16-byte PC room DL with raw
((u8*)gdl)[k] / ((u32*)gdl)[i] 8-byte-Gfx byte/word indexing. It runs on
every shot that resolves against background geometry (bgTestBulletHitBackground),
so it AVs the first time you shoot a wall or floor in a real playthrough --
invisible to the no-input level sweep.
Ported #ifdef PORT (N64 path kept verbatim under #else):
- vtxoff / vtxbase / op read through the .dma view, matching the
already-ported bgBuildRoomVtxBounds.
- G_TRI1 vertex indices from (w1 >> 16/8/0) & 0xff, /10.
- the 12 G_TRI4 index nibbles recovered from (u32)gdl->words.w0/.w1,
each mapped from the N64 byte/halfword layout (table in the code
comment; cross-checked twice).
- both texturenum recoveries -> -1: the KSEG0 *(u16*)(w1-8) deref is not
valid for the port's converted GDLs (same as D135); texnum only selects
the bullet-impact decal/sound, parked with D77.
Build: green (ntsc-final). NOT VERIFIED -- a no-input GE_PCDUMP capture
never fires a bullet so it never enters this function; needs a firefight
into a wall on an idle machine. This session's box was too thrashed to get
a clean run (D153 driver-crash-under-load). Committed as WIP so the port
work isn't lost; next session: idle-machine wall-shoot test + -level_09
framediff, then mark verified.
§F/HANDOFF D154, PORT-LEARNINGS §B (D135 corollary).
D150 (src/str.c) — NULL-tolerant strcpy/strncpy/strcat under #ifdef PORT.
The watch BRIEF/OBJECTIVES pages build text with strcat(buf, langGet(id));
langGet() returns NULL on PC for a string bank the in-level flow never
loaded, and str.c's primitives are __nonnull__ builtins so a plain param
NULL check is optimised away — the guard launders the ptr through an empty
asm (GE_IS_NULL). NULL src -> empty string instead of an AV crash.
D151 (src/bondtypes.h) — watch data (objectives / mission background / M /
Q / Moneypenny) all rendered blank. propDef types 35 (WatchMenuObjectiveText)
and 23 (ObjectiveStart) store a plain s32 lang slot id in their 3rd word;
the decomp structs decode it as [u16 reserved][u16 text] reading text at
offset 0xA, which only works because the id is in the low 16 bits of a
big-endian word. d88_propdefs.py bswap32's the whole word -> text @0xA
reads 0 -> langGet(0) -> NULL. Fix: text is a full u32 at the word offset
under #ifdef PORT (N64 layout kept under #else). No converter change, no
sidecar regen. This is the D143/D150 "briefing text renders blank" root cause.
D152 (port/src/libultra.c) — mission-failed / objective-failed permanent
black screen. The D147 recursive-mutex behind osSetIntMask can be leaked
(libaudio has unbalanced / early-return mask calls; a transient thread can
acquire OS_IM_NONE and exit) -> mainThread + amMain both block forever in
alEvtq* -> gfx pipeline stalls. Live gdb.txt from the user's session
confirmed the deadlock (mainThread in the mission-fail audio fade-out
sndSetScalerApplyVolumeAllSfxSlot -> alEvtqPostEvent). Mitigation: model
the mask as a self-healing logical lock (owner/depth under s_imMx + condvar)
— normal microsecond critical sections behave as before, but a waiter
blocked > 2s concludes the holder leaked it, logs the stale owner + the
stealing caller's return address, and steals the section. Worst case is a
~2s audio hiccup + a log line naming the leak, instead of a hung game.
Proper narrow fix (dedicated ALEventQueue lock) still owed.
D153 investigated and dismissed — the "-level_09 crashes at frame ~900"
seen while regression-testing was D117/D134 host-scheduling flakiness under
heavy concurrent build+run load; on a quiet machine -level_09 runs past
frame 2400 clean with all three changes applied.
Build: green (ntsc-final). Golden framediff not run clean (same machine-load
flakiness); one quiet -level_09 run = 2400+ frames, 0 heartbeats.
James Dansereau