vgsp: m171 event completion fixes (minimal cudaEventSynchronize repro now passes) + AGENT-F R2 write/read consistency ledger

- gsp_rm_emulation: the event burst budget becomes a time-windowed rate
  (1 s) with completion posts allowed headroom over hint re-posts; the
  periodic hint dedups events posted within 250 ms (strictly below its own
  500 ms cadence - an equal window races the boundary and skips forever);
  force_post skips retired slots (rc=22 spam was an artifact). Verified:
  the minimal cudaEventCreate/Record/Synchronize repro and a second repro
  with cross-stream waits and elapsed-time now complete on a live boot;
  previously cudaEventSynchronize hung forever.
- gsp_sema_service: write/read consistency ledger (AGENT-F R2). Large
  device-destination writes record {va_page -> pa, aperture}; the launch
  data fill and device-source reads prefer the recorded page whenever
  their own resolution disagrees (sema-ledger DIVERGE log). Session
  scoped, zero cost while empty.

1 files changed, 3 insertions, 1 deletions

diff --git a/kernel/gsp_rm_emulation.h b/kernel/gsp_rm_emulation.h
index e464809..96ea16a 100644
--- a/kernel/gsp_rm_emulation.h
+++ b/kernel/gsp_rm_emulation.h
@@ -197,11 +197,13 @@ struct gsp_rm_emu {
 		u32	notify_index;
 		u32	data;
 		bool	valid;
+		unsigned long last_post;	/* m171: jiffies of the last posted record */
 	} events[GSPRM_MAX_EVENTS];
 	u32		events_n;
 	u64		events_posted;
 	u64		events_dropped;
-	u32		event_burst;	/* event records written since the last reply */
+	u32		event_burst;	/* event records written in the current window */
+	unsigned long event_burst_at;	/* m171: burst window start (time-based reset) */
 };
 
 int  gsp_rm_emu_init(struct nv_vgpu_device *vgpu, struct gsp_emu_context *ctx);