[$] The 7.0 scheduler regression that wasn't

One of the more significant changes in the 7.0 kernel release is to use the lazy-preemption mode by default in the CPUscheduler. The scheduler developers have wanted to reduce the number ofpreemption modes for years, and lazy preemption looks like a step towardthat goal. But then there came this reportfrom Salvatore Dipietro that lazy preemption caused a 50% performanceregression on a PostgreSQL benchmark. Investigation showed that thesituation is not actually so grave, but the episode highlights just howsensitive some workloads can be to configuration changes; there may besurprises in store for other users as well.