Hi, I didn't feel that benchmarks were necessary in this case, since the result is clearly visible: either it does or does not swap under a given workload. We

Hi, I didn't feel that benchmarks were necessary in this case, since the result is clearly visible: either it does or does not swap under a given workload. We did run benchmarks, but only to prove that the performance was nominal with and without the setting in place, swap behavior aside, to ensure that this doesn't introduce some regression. Regards, Jeremy