Re: [PATCH 3/4] x86/nmi: wait for all CPUs in check_nmi_watchdog()

[David Vrabel <david.vrabel@citrix.com>]

On 14/05/14 14:59, Jan Beulich wrote:
>>>> On 14.05.14 at 14:58, <david.vrabel@citrix.com> wrote:
>> The counting of a CPUs NMIs in check_nmi_watchdog() is only reliable
>> if all CPUs have been spinning for 5 or more ticks.  There may be
>> delays in waking other CPUs from deep power states that can mean that
>> when the counts are checked CPUs haven't run for long enough.
> 
> 5 ticks ought to be a couple of orders of a magnitude longer than
> the worst possible wakeup time. I.e. I don't buy this argument
> without actual numbers to support it.

This was a change Andrew asked for and the reasoning he used, but I
agree that it seems highly implausible.

I think this is a worthwhile cleanup anyway so you can apply with the
below commit message or drop it (your preference).

David

8<-------
x86/nmi: wait for all CPUs in check_nmi_watchdog()

The counting of a CPUs NMIs in check_nmi_watchdog() is only reliable
if all CPUs have been spinning for 5 or more ticks.  Whilst its highly
implausible that this won't happen in practise, explicitly wait so it's
clear that this is required.