handoff: VSync/event-wedge fixes + iterate 2.A–2.BC research notes

Source changes (dormant parity infra, retained from iterate 2.AI/2.AO):
- xenia-kernel/exports.rs: nt_create_event manual_reset polarity +
  related event wiring
- xenia-gpu/mmio_region.rs: D1MODE_VBLANK_VLINE_STATUS hardcode parity

Also lands the audit-runs/ analysis notes (.md/.txt/.json digests) for the
iterate 2.x VSync/0x10e8/0x1004 wedge investigation. Raw trace dumps
(.jsonl/.gz/.csv/.stdout) and agent worktrees (.claude/) are gitignored as
regenerable local artifacts — see memory + HANDOFF for the running findings.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

audit-runs/phase-c23-scheduler-determinism-plan/probes/jitter_profile.py

@@ -0,0 +1,97 @@
+#!/usr/bin/env python3
+"""Phase C+23 jitter profile probe.
+
+Reads canary jsonls (jitter-1/2/3 + cold-c21 + any fresh runs) and extracts:
+- total tid=6 events seen within the first ~120k indices
+- the exact event sequence on tid=6 around idx [104,595..104,620]
+- count of wait.begin events on tid=6 by SID
+- count of contention-prone events (wait.begin, kernel.call RtlEnter / RtlLeave)
+
+Designed to stream line-by-line and not load multi-GB jsonls into RAM.
+"""
+
+import json
+import os
+import sys
+from collections import Counter, defaultdict
+
+WINDOW_LO = 104_595
+WINDOW_HI = 104_620
+TID = 6
+TID_EVENT_LIMIT = 120_000
+
+
+def profile(path: str):
+    if not os.path.exists(path):
+        return None
+    tid6_events = 0
+    waitbegins = Counter()
+    importcalls = Counter()
+    kernelcalls = Counter()
+    window_events = []
+    tid_idx = -1
+
+    with open(path, "rb") as fh:
+        for raw in fh:
+            # Cheap reject before json parse: must contain `"tid":6,`
+            if b'"tid":6,' not in raw and b'"tid": 6,' not in raw:
+                continue
+            try:
+                ev = json.loads(raw)
+            except Exception:
+                continue
+            if ev.get("tid") != TID:
+                continue
+            tid_idx = ev.get("tid_event_idx", tid_idx + 1)
+            tid6_events += 1
+            kind = ev.get("kind", "")
+            if kind == "wait.begin":
+                sids = ev.get("payload", {}).get("handles_semantic_ids") or []
+                for s in sids:
+                    waitbegins[s] += 1
+            elif kind == "import.call":
+                name = ev.get("payload", {}).get("name", "")
+                importcalls[name] += 1
+            elif kind == "kernel.call":
+                name = ev.get("payload", {}).get("name", "")
+                kernelcalls[name] += 1
+
+            if WINDOW_LO <= tid_idx <= WINDOW_HI:
+                summary = {
+                    "idx": tid_idx,
+                    "kind": kind,
+                    "name": ev.get("payload", {}).get("name", ""),
+                }
+                if kind == "wait.begin":
+                    summary["sid"] = (ev.get("payload", {}).get("handles_semantic_ids") or [None])[0]
+                    summary["timeout_ns"] = ev.get("payload", {}).get("timeout_ns")
+                window_events.append(summary)
+
+            if tid_idx > TID_EVENT_LIMIT:
+                break
+
+    return {
+        "path": path,
+        "tid6_total_seen": tid6_events,
+        "waitbegins_by_sid": dict(waitbegins.most_common(10)),
+        "rtlenter_calls": importcalls.get("RtlEnterCriticalSection", 0),
+        "rtlleave_calls": importcalls.get("RtlLeaveCriticalSection", 0),
+        "window_events": window_events,
+    }
+
+
+def main(paths):
+    out = {}
+    for p in paths:
+        print(f"profiling {p}...", file=sys.stderr)
+        r = profile(p)
+        if r is None:
+            print(f"  (missing)", file=sys.stderr)
+            continue
+        out[os.path.basename(p)] = r
+    json.dump(out, sys.stdout, indent=2)
+    print()
+
+
+if __name__ == "__main__":
+    main(sys.argv[1:])