On 13/04/2021 10:12, liuqi (BA) wrote:
I do wonder why we even need maintain pcie_pmu->cpumask Can't we just use cpu_online_mask as appropiate instead??Sorry, missed it yesterday.It seems that cpumask is always same as cpu_online_mask, So do we need to reserve the cpumask sysfs interface?
I'm not saying that we don't require the cpumask sysfs interface. I am just asking why you maintain a separate cpumask, when, as I said, they seem the same.
Thanks, John
