On 2018-09-30 11:16:41 [+0800], kernel test robot wrote:
FYI, we noticed a -2.4% regression of will-it-scale.per_process_ops due to commit:
commit: d884021f52609407c7943705b3e54b1642fa10cb ("[PATCH 1/2] mm/swap: Add pagevec
-2.4% regression reads to me like an improvement of 2.4%. Which is odd
because I wouldn't expect a change in behaviour from 1/2. 2/2 yes, under
certain workloads but not from 1/2.