It would be nice if the profile showed the selfcpu spent while world.tick_usage is over 100%. Depending on how profiling is done, this should be as simple as just checking when the tick started (iirc you have a global var for that) and doing the math¹ on that when it saves/increments the self_cpu info.

The extra load /tg/ (and all of ss13) has gotten from the Sseth video about ss13 has made finding these tick overrunning procs even more necessary now that most of our servers are hitting triple digits player counts.

This kind of change shouldn't (hopefully) require a major version bump.

_{¹tick_overrun_ms += max(proc_self_cost_this_run + proc_start_ms - tick_start_ms - tick_lag_ms, 0)}