I agree that it is wise to intersect cq_cpumask with online_cpumask in cq_cpulist_set(), and if cpuset_weight(cq_cpumask & online_cpumask) == 0 -- ignore cq_cpumask and stick with default behavior.
I don't think cpu offlining is problematic here. If CPU is offlined -- the IRQ (and queue) will be served by another online CPU. Yes, it will break IRQ affinity and make cq_cpumask setting pointless, but this is system administrator's problem. The parameter is only useful for fine tuning and some resource planning in advance is required. I don't think that this side effect is enough to motivate "unsafe" tag. -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1831566 Title: [realtime app] not possible to redirect drivers/nvme IRQs from realtime cpuset Status in linux package in Ubuntu: Confirmed Bug description: We're running realtime application on Ubuntu 16.04 with linux-image 4.15 and found it impossible to get rid of jitter introduced by Intel NVMe IRQs. I'm providing here a patch which solved the issue for us. The realtime application is bound to isolated CPUs (one thread per CPU, nohz_full= in kernel cmdline, all IRQs moved to housekeeping CPUs), application doesn't use any linux kernel syscalls except in startup phase so we don't expect any interruptions of the application from the kernel or HW. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1831566/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
