Re: [PATCH] printk/nmi: restore printk_func in nmi_panic

[Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>]

On (02/29/16 11:31), Petr Mladek wrote:
> > > > diff --git a/include/linux/kernel.h b/include/linux/kernel.h
> > > > index f4fa2b2..3ee33d5 100644
> > > > --- a/include/linux/kernel.h
> > > > +++ b/include/linux/kernel.h
> > > > @@ -469,10 +469,12 @@ do {									\
> > > >  	cpu = raw_smp_processor_id();					\
> > > >  	old_cpu = atomic_cmpxchg(&panic_cpu, PANIC_CPU_INVALID, cpu);	\
> > > >  									\
> > > > -	if (old_cpu == PANIC_CPU_INVALID)				\
> > > > +	if (old_cpu == PANIC_CPU_INVALID) {				\
> > > > +		printk_nmi_exit();					\
> > > 
> > > This might end up in a deadlock that printk_nmi() wanted to avoid.
> > 
> > aha, I see.
> > 
> > > I think about a compromise. We should try to get the messages
> > > out only when kdump is not enabled.
> > 
> > can we zap_locks() if we are on nmi_panic()->panic()->console_flush_on_panic() path?
> 
> That is the problem. zap_locks() is not a solution.
> 
> First, it handles only lockbuf_lock and console_sem. There are other
> locks used by particular consoles that might cause a deadlock.

yes, well, that's true for panic() in general. we can't take care of
all of the locks that possibly could have been in busy state at the
time CPU received STOP_IPI or entered panic(). we can't even safely
iterate all of the consoles and call ->reset() on them (provided that
such struct console callback will ever be implemented) because
smp_send_stop() is free to leave some CPUs online.


> Second, re-initializing locks is dangerous of its own. If they are
> released by some other CPU that is still running, you might end up
> in a deadlock because of a double release. In fact, I think that it
> actually increases the risk. If there are more than 2 CPUs than
> it is more likely that a printk is running on another CPU than
> on the current one.

panic calls debug_locks_off(), so chances *seem* to be lower.

hm... we are (sort of) on the safe side if we know that smp_send_stop() has
stopped all the CPUs but panic cpu; so zap_locks() is safe (to some extent of
course) when we know that num_online_cpus() == 1 in console_flush_on_panic().

	-ss