On Thursday 19 Nov 2020 at 20:39:07 (+0000), Will Deacon wrote: > Ah, so in doing this I realised I don't like arch_cpu_possible_mask() so > much because it makes it sound like a back-end to cpu_possible_mask, but > the two are really different things. > > arch_task_cpu_possible_mask() might work? Yes, making it explicit in the name that this is a task-specific thing doesn't hurt. Thanks, Quentin