On 06/24, Oleg Nesterov wrote: > > So we need percpu_down_write_dont_block_readers(). I already thought > about this before, I'll try to make the patch tomorrow on top of your > changes.
Never say tomorrow... > This means that we do not need task_struct->cpuhp_ref, but we can't > avoid livelock we currently have: cpu_hotplug_begin() can never succeed > if the new readers come fast enough. Like with any other "recursive" lock. Peter, I know you don't like the 1st patch. And yes, we could add another mutex into percpu_rw_semaphore instead. But I think it would be better to rely on rcu_sync_enter(). As for completion, we can remove it later. Nevermind, the actual change is 3/3 and it looks simple. Oleg. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/