Messages in this thread | | | Date | Wed, 12 Oct 2011 18:33:59 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH 1/3] x86, reboot: Use NMI instead of REBOOT_VECTOR to stop cpus |
| |
* Don Zickus <dzickus@redhat.com> wrote:
> On Wed, Oct 12, 2011 at 09:30:25AM +0200, Ingo Molnar wrote: > > > > * Don Zickus <dzickus@redhat.com> wrote: > > > > > + if (register_nmi_handler(NMI_LOCAL, smp_stop_nmi_callback, > > > + NMI_FLAG_FIRST, "smp_stop")) > > > + return; /* return what? */ > > > > That comment looks a bit odd. > > Yeah, I copied it from the kdump code because it seemed relevant. > The point was to express the paranoid concern, if we can't register > the NMI handler for whatever reason, what happens!?. How do we > explain to anyone we failed to shut down the other cpus? > > I can expand the comment to be more specific in the paranoia. I > just wasn't sure the right way to handle that failure case.
Just add something like:
/* Note: we ignore failures here */
if there's nothing intelligent possible.
Thanks,
Ingo
| |