Messages in this thread | | | Subject | Re: [PATCH] printk: robustify printk | From | Andi Kleen <> | Date | Fri, 08 Aug 2008 23:35:20 +0200 |
| |
Peter Zijlstra <a.p.zijlstra@chello.nl> writes:
Hopefully nobody puts a printk into call_rcu now... Also I'm not sure if there wouldn't be livelocks if someone puts printk into some other parts of RCU (e.g. each RCU adding more RCU events -- haven't checked that in detail so I might be wrong)
So you just trade one "hands off" part to another.
> + kws = &__get_cpu_var(kws); > + if (!kws->pending) { > + kws->pending = 1; > + call_rcu(&kws->head, __wake_up_klogd);
How about a big comment describing why you do this? And you should add comments to all the new hands-off parts where printk is not allowed now.
-Andi
| |