Re: [PATCH -v3] perf, x86: try to handle unknown nmis with running perfctrs

From: Cyrill Gorcunov <gorcunov@gmail.com>
To: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@elte.hu>, Don Zickus <dzickus@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Lin Ming <ming.m.lin@intel.com>,
	"fweisbec@gmail.com" <fweisbec@gmail.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"Huang, Ying" <ying.huang@intel.com>,
	Yinghai Lu <yinghai@kernel.org>, Andi Kleen <andi@firstfloor.org>
Subject: Re: [PATCH -v3] perf, x86: try to handle unknown nmis with running perfctrs
Date: Tue, 24 Aug 2010 21:21:57 +0400	[thread overview]
Message-ID: <20100824172157.GE5723@lenovo> (raw)
In-Reply-To: <20100824172022.GD5723@lenovo>

On Tue, Aug 24, 2010 at 09:20:22PM +0400, Cyrill Gorcunov wrote:
> On Tue, Aug 24, 2010 at 07:09:36PM +0200, Robert Richter wrote:
> > On 24.08.10 12:22:52, Cyrill Gorcunov wrote:
> > > Btw, guys, I fail to see how new nmi_watchdog work, we have
> > > default_do_nmi
> > >   if (!(reason & 0xc0)) {
> > >     if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == == NOTIFY_STOP)
> > >       return
> > >     if (nmi_watchdog_tick(regs, reason))
> > >       return
> > > 
> > > but perf_event_nmi_handler returns NOTIFY_STOP when watchdog is perf event
> > > and nmi_watchdog_tick _never_ called, or (most probably) I miss something?
> > 
> > The watchdog is disabled during profiling (perf and oprofile) by
> > calling disable_lapic_nmi_watchdog().
> > 
> > -Robert
> >
> 
> Huh? iirc Don have switched nmi watchdog to native perf subsystem, ie watchdog
> uses PERF_COUNT_HW_CPU_CYCLES event, letme check...
>  
> 	-- Cyrill

False alarm, perf watchdog uses own handler. Sorry for noise ;)

	-- Cyrill