On Tue, Oct 20, 2020= at 8:46 PM Matt Spinler <mspi= nler@linux.ibm.com> wrote:

On 10/19/2020 10:23 AM, Thu Ba Nguyen wrote:
> Thanks for your reply Matt Spinler, Can you show me the discussion > threads? I also...
> This Message Is From an External Sender
> This message came from outside your organization.
>
> Thanks for your reply Matt Spinler,
>
> Can you show me the discussion threads?

Sure: https://lists.ozlabs.org/pi= permail/openbmc/2019-October/018967.html

>
> I also thought about the solution for that features:
> In the current hwmon we support GPIOCHIP=C2=A0+ GPIO=C2=A0option which= used to
> enable sensors to read. In the hwmon code, we just set that pin and > wait before reading.
> I think we can support a similar option named GPIOENABLE=C2=A0+ GPIOV.= When
> the status of Gpio pin defind in GPIOEANBLE=C2=A0match with GPIOV.
> That sensors will be read and update to Dbus.
> If not it will be removed from DBus until the GPIO pin math GPIOV.
> Maybe we can have many different solutions.

As Ed mentioned, I think a good direction to start with is how
dbus-sensors handles it, so we=C2=A0 can have
common behavior.=C2=A0 I believe they look at the host state D-Bus property=
and still keep the sensor
on D-Bus even when power is off.

=C2=A0

[Thu= ] Yah, I think =C2=A0using=C2=A0the host state D-Bus property=C2=A0to verify host state.

Then handling the host sensors based on this status is a better solution.<= /span>

Keeping the sensors on= Dbus can cause the value or the status of warning/critical value of sensor= s

is out update. I exp= ected that the host sensors will be removed from Dbus when the host is off.=

>
> If you don't mind, can you tell me how IBM supports that features?=

We lucked out out in that the driver was only loaded when power was on.
=

[Thu] What will be displayed when you open health --> = sensors page in OpenBmc web?=C2=A0

Are the host sensors still the= re and value is "na" or some things like that?

>
> Regards.
> Thu Nguyen.
>
> On Mon, Oct 19, 2020 at 9:16 PM Matt Spinler <mspinler@linux.ibm.com
> <mailto:mspinler@linux.ibm.com>> wrote:
>
>
>
>=C2=A0 =C2=A0 =C2=A0On 10/18/2020 8:58 AM, Thu Ba Nguyen wrote:
>=C2=A0 =C2=A0 =C2=A0> Dear, I'm supporting the host sensors for = Ampere Computing LLC
>=C2=A0 =C2=A0 =C2=A0> platform. We are...
>=C2=A0 =C2=A0 =C2=A0> This Message Is From an External Sender
>=C2=A0 =C2=A0 =C2=A0> This message came from outside your organizati= on.
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> Dear,
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> I'm supporting the host sensors for Ampere= Computing LLC platform.
>=C2=A0 =C2=A0 =C2=A0> We are using phosphor-hwmon to update values o= f sensors and
>=C2=A0 =C2=A0 =C2=A0monitoring
>=C2=A0 =C2=A0 =C2=A0> sensors warning/errors base on threshold setti= ng.
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> There are some sensors which are turned off wh= en host=C2=A0Off. It
>=C2=A0 =C2=A0 =C2=A0can be
>=C2=A0 =C2=A0 =C2=A0> the sensors reported by host or voltage/temper= ature/power sensors
>=C2=A0 =C2=A0 =C2=A0> =C2=A0which use the same power source with hos= t.
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> I researched in openBmc sensor-architecture do= cuments=C2=A0but can't
>=C2=A0 =C2=A0 =C2=A0find
>=C2=A0 =C2=A0 =C2=A0> any option to enable/disable sensors base on o= ne status or GPIO
>=C2=A0 =C2=A0 =C2=A0pins.
>=C2=A0 =C2=A0 =C2=A0> I can't use REMOVERCS.
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> Research in phosphor-hwmon code, I don't s= ee the answer too.
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> Do we have any options/solution to=C2=A0Enable= /Disable some sensors when
>=C2=A0 =C2=A0 =C2=A0> Host On/Off?
>
>=C2=A0 =C2=A0 =C2=A0Hi,
>=C2=A0 =C2=A0 =C2=A0The phosphor-hwmon code doesn't support that ye= t.=C2=A0 It has been
>=C2=A0 =C2=A0 =C2=A0discussed
>=C2=A0 =C2=A0 =C2=A0before but nobody
>=C2=A0 =C2=A0 =C2=A0has implemented it.
>
>=C2=A0 =C2=A0 =C2=A0>
>=C2=A0 =C2=A0 =C2=A0> Thanks.
>=C2=A0 =C2=A0 =C2=A0> Thu Nguyen.
>

--0000000000001dad3e05b2226c28--