pm/testing baseline: 48 runs, 3 regressions (v6.5-rc1-21-g3c61a03588dd8)