Messages in this thread | | | Date | Thu, 13 Dec 2012 08:32:33 -0800 | From | Tejun Heo <> | Subject | Re: [RFC PATCH v4 1/9] CPU hotplug: Provide APIs to prevent CPU offline from atomic context |
| |
Hello, Oleg.
On Thu, Dec 13, 2012 at 05:17:09PM +0100, Oleg Nesterov wrote: > Hmm. I thought that __this_cpu_* must be safe under preempt_disable(). > IOW, I thought that, say, this_cpu_inc() is "equal" to preempt_disable + > __this_cpu_inc() correctness-wise.
this_cpu_inc() equals local_irq_save() + __this_cpu_inc().
> And. I thought that this_cpu_inc() is safe wrt interrupt, like local_t.
Yes, it is safe.
> But when I try to read the comments percpu.h, I am starting to think that > even this_cpu_inc() is not safe if irq handler can do the same? > > Confused...
Yeah, the comment is confusing and the way these macros are defined doesn't help. There used to be three variants and it looks like we didn't update the comment while removing the preempt safe ones. Gotta clean those up. Anyways, yes, this_cpu_*() are safe against irqs.
Thanks.
-- tejun
| |