ok, then I am done here. Bye.
Seems interesting. Since the memory is never released but the leaking process is "just" closed down I presume that eventually the memory will still get full if you play enough? I guess that depends on how you detect the memory leak.
The script requests the size of the nonPagedPool every second (by default) from the system.
It then calculates the difference between the current and previous messurement (delta-value). If this delta gets up to 10MByte or more (by default) the programm "assumes" that there is a memory-leak and kills the bf4-process.
I've tested alot with the size of this delta-limit and my system was always leaking memory when the 10 MByte-Cap was reached. So this assumption might not be 100% accurate - but nearly 😉
As a "fallback" there is second limit: When the general size of the NP-Pool reaches 500MByte (by default) then the program also assumes that there must be a memoryleak and kills the BF4-Process.
This fallback prevents the system from out-of-memory, when the delta-limit is set too high.
To your question: Yes, the NP-Pool will still raise with this workaround activated - but much less than without it 😉
As an example: Normally when I experienced a memory-leak on my pc while playing bf4, the NP-Pool raised up to 6 GByte (75% of my total memory) until I realized it.
No with the script running in the background, I experienced 3 memory-leaks, but my NP-Pool only raised to 270MB without restarting my pc!
But to make this clear: This is not meant to be a total solution for this issue - it's more like a firefighter (just "extinguishing the fire") 😉
