Re: power-off delay/hang due to commit 6d25be57 (mainline)

From: "Rafael J. Wysocki" <rafael@kernel.org>
To: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>,
	Zhang Rui <rui.zhang@intel.com>,
	Robert Moore <robert.moore@intel.com>,
	Erik Kaneda <erik.kaneda@intel.com>,
	"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
	Len Brown <lenb@kernel.org>, Thomas Gleixner <tglx@linutronix.de>,
	Peter Zijlstra <peterz@infradead.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
	Stephen Berman <stephen.berman@gmx.net>,
	"open list:ACPI COMPONENT ARCHITECTURE (ACPICA)"
	<devel@acpica.org>
Subject: Re: power-off delay/hang due to commit 6d25be57 (mainline)
Date: Tue, 14 Jul 2020 17:53:15 +0200	[thread overview]
Message-ID: <CAJZ5v0iogereapmhcFi=iXHsjnzmC26mewUSTY3+5O3ei5kfDQ@mail.gmail.com> (raw)
In-Reply-To: <20200714141135.47adndrsdgpiqfy4@linutronix.de>

On Tue, Jul 14, 2020 at 4:11 PM Sebastian Andrzej Siewior
<bigeasy@linutronix.de> wrote:
>
> On 2020-07-14 15:54:57 [+0200], Rafael J. Wysocki wrote:
> > On Tue, Jul 14, 2020 at 3:44 PM Sebastian Andrzej Siewior
> > <bigeasy@linutronix.de> wrote:>
> > > On 2020-06-24 23:49:52 [+0200], Stephen Berman wrote:
> > >
> > > Let me summarize the thread here:
> > >
> > > On Stephen's system, ACPI informs the thermal zone driver to poll the
> > > temperature every second and the driver does so.
> > > The driver queries the temperature by invoking acpi_evaluate_integer()
> > > which invokes (at some point) acpi_ev_queue_notify_request().
> >
> > Well, I don't quite see how acpi_ev_queue_notify_request() can be
> > invoked from the acpi_evaluate_integer() code path.
> >
> > Do you have a call trace showing that?
>
> So the trace in
>      https://lore.kernel.org/linux-acpi/87o8pjh1i0.fsf@gmx.net/
>
> shows the pattern and we nailed it down that it comes from
> thermal_get_temp().

acpi_evaluate_integer() doesn't show up in the trace, though, AFAICS.

> I assumed acpi_ex_opcode_2A_0T_0R() since the other
> candidate was acpi_ev_asynch_execute_gpe_method().

Which probably is the case.  Specifically

acpi_ev_asynch_execute_gpe_method: Evaluate _L66

is likely to cause the Notify() to be dispatched.

> Stephen, the patch attached adds a WARN_ON() statement which will
> produce a stack trace (4 or so). Could please run 'dmesg' after a while
> and send it back. There should be a
>     "WARNING in drivers/acpi/acpica/evmisc.c"
>
> statement or something along the lines.
>
> Rafael, are you also interested in an ACPI dump?

That might help a bit.

So what probably happens is that poking at the TZ causes a GPE to
trigger and a Notify() to be dispatched which then goes into the
workqueue for execution.

Now, I'm not sure what happens to those Notify() items, though.  They
each should cause a handler (in the thermal driver) to be executed,
but does that happen?