On Thu 2019-03-07 10:53:48, John Ogness wrote: > On 2019-03-04, Sergey Senozhatsky <sergey.senozhatsky.work@xxxxxxxxx> wrote: > > If there are setups which can be fully !atomic (in terms of console > > output) then we, essentially, have a fully preemptible kthread printk > > implementation. > > Correct. I've mentioned in another response[0] some ideas about what > could be done to aid this. > > I understand that fully preemptible kthread printing is unacceptable for > you. Since all current console drivers are already irq safe, I'm > wondering if using irq_work to handle the emergency printing for console > drivers without write_atomic() would help. (If the printk caller is in a > context that write() supports, then write() could be called directly.) > This would also demand that the irq-safe requirements for write() are > not relaxed. The printk-kthread might still be faster than irq_work, but > it might increase reliability if an irq_work is triggered as an extra > precaution. It is getting more and more complicated. The messages would be pushed directly, from irq, and kthread. It would depend how the code would look like but I am not much optimistic. Note that you could not pass any data to the irq_work handler. It would need to iterate over the logbuffer and take care of all non-handled emergency messages. Anyway, we could solve this later. We need to keep the current console_unlock() handling as a fallback until enough consoles support the direct mode anyway. Best Regards, Petr
