Alex Polite wrote: > And with CONFIG_PREEMPT=y My understanding of CONFIG_PREEMPT=y is that it does not do what people think it does. Last time I read the google hits it was considered good for simulating MP on single cpu systems and finding bugs but not so good for reducing latency. Unless you have some recent benchmark data which shows a real reduction in latency I would stick with CONFIG_PREEMPT=n. Does anyone have any real data that shows the utility of CONFIG_PREEMPT=y over CONFIG_PREEMPT=n?
Bob
signature.asc
Description: Digital signature
