On (03/13/19 09:40), Sebastian Siewior wrote: > On 2019-03-13 09:19:32 [+0100], John Ogness wrote: > > recursive situation. As you are pointing out, the notification/wake > > component of printk_safe will still be needed. I will leave that (small) > > part in printk_safe.c. > > Does this mean we keep irq_work part or we bury it and solve it by other > means? That's a very good question. Because if we add console_trylock() to printk(), then we can't invoke ->atomic() consoles when console_sem is already locked, because one of the registered drivers is currently being modified by a 3rd party and printk(), thus, must stay away. Once that modification will be done console_unlock() will print all pending messages. This is current design. And this conflicts with the whole idea of ->atomic() consoles. So may be we need a whole new scheme in this department as well. For instance [*and this is completely untested idea* !!!] *May be* we can take a closer look and find cases when ->atomic consoles don't really depend on console_sem. And *may be* we can split the console drivers list and somehow forbid removal and modification (ioctl) of ->atomic consoles under us. Assuming that this is doable we then can start iterating ->atomic consoles list with unlocked console_sem. Non->atomic consoles or consoles which depend on console_sem (VT, fbcon and so on) will stay in another list, and we will take console_sem before we iterate that list and invoke those drivers. One more time - a completely random thought. -ss
