Emulated images keep a running total of their decoded and ETC2/EAC sizes over every mip and layer. --profile writes both as counters, and summarize_profile_trace.py prints their peaks.
vr.frame_rate_limit / MOCKTAIL_VR_FRAME_RATE_LIMIT replaces
graphics.frame_rate_limit while VR is active and takes the same values.
It defaults to unlimited so xrWaitFrame paces Roblox at the headset
refresh rate instead of Roblox's own 60 FPS cap.
MOCKTAIL_VR_RENDER_SCALE scales the runtime's recommended eye extent.
SteamVR's recommendation for a Quest 2 at 90 Hz (2244x2352 per eye) missed
the 11.1 ms budget and halved the rate to 45 FPS.
MOCKTAIL_VR_UNPACED=1 stops guest frames from blocking in xrWaitFrame:
the open XR frame spans several guest frames and is submitted once per
display period, for measuring maximum render rate.
Uploads decode on a background dispatcher and the submit carries a
timeline semaphore wait, so the GPU waits for the texture rather than
the render thread. The gather of the application's compressed bytes
moves there too when they are already mapped; unmapping or freeing that
memory drains the decodes still reading it.
Staged copies gain an explicit host-write barrier. A queue submission
only makes host writes from before it visible to the device, and these
writes land after the submit.
Devices without timelineSemaphore, and submits whose pNext already sizes
arrays by their semaphore counts, decode inline as before.
Blade Ball, median of three: worst vkQueueSubmit 140ms -> 1ms, p99
4.9ms -> 0.05ms, total 2534ms -> 309ms. Frame interval unchanged; its
tail is the engine's own main-thread step.
Release the GameMode request when the daemon narrows the CPU mask; the pin
concentrates Roblox's worker pool and is re-applied every few seconds, so
ending the request is the only thing that lifts it. `gamemode: on` keeps it.
Emit texture uploads on worker threads instead of serially inside
vkQueueSubmit, in two passes so mip 0 lands before the levels scaled from it.
Decode ETC2 through a persistent pool and lower the parallel floor to one
128x128 image, so small batches stop running single-threaded.
Document all three in docs/PERFORMANCE.md.
vii