Re: [lm-sensors] [PATCHv4 1/1] Hwmon: Add core/pkg Threshold

[Guenter Roeck <guenter.roeck@ericsson.com>]

Hi Durga,

On Mon, Sep 19, 2011 at 11:27:56PM -0400, R, Durgadoss wrote:
> Hi Guenter,
> 
> > -----Original Message-----
> > From: Guenter Roeck [mailto:guenter.roeck@ericsson.com]
> > Sent: Monday, September 19, 2011 10:54 PM
> > To: Jean Delvare
> > Cc: R, Durgadoss; lm-sensors@lm-sensors.org
> > Subject: Re: [lm-sensors] [PATCHv4 1/1] Hwmon: Add core/pkg Threshold Support
> > to Coretemp
> > 
> > On Sun, 2011-09-18 at 13:24 -0400, Jean Delvare wrote:
> > > On Sat, 17 Sep 2011 11:09:43 -0700, Guenter Roeck wrote:
> > > > Hi Durga,
> > > >
> > > > On Sat, Sep 17, 2011 at 01:40:41PM -0400, R, Durgadoss wrote:
> > > > > Hi Guenter,
> > > > >
> > > > [ ... ]
> > > >
> > > > > > > For me, it looks like, we need not know whether the threshold is
> > upper or
> > > > > > lower.
> > > > > > > Anyway, for every threshold, we will get two interrupts (for either
> > > > > > direction)
> > > > > > > So, the user space can assume either a lower threshold and look for 0
> > in the
> > > > > > > Corresponding alarm interface Or a higher threshold and look for 1 in
> > the
> > > > > > > alarm interface. Will this not work ?
> > > > > > >
> > > > > > For a lower threshold, "alarm" implies "temperature is at or below
> > threshold",
> > > > > > In other words, "alarm" can mean that a value is above or below a given
> > > > > > threshold -
> > > > > > it has a semantics that depends on its context.
> > > > > >
> > > > > > This context is not known in the case of a generic threshold. This is
> > why I
> > > > > > suggested
> > > > > > to use a more neutral term, such as "triggered", which would imply "at
> > or above
> > > > > > threshold" (or possibly just "threshold triggered" if the direction is
> > not
> > > > > > known)
> > > > > > without attaching a semantics to it.
> > > > > >
> > > > >
> > > > > Ok. I agree. We can use tempX_thresholdY_triggered.
> > > > >
> > > > I'd like to hear Jean's opinion first.
> > >
> > > I'm fine with your proposal, yes.
> > >
> > > > Also, if we introduce new attributes, we should probably reinstate the old
> > "max".
> > >
> > > Definitely. And preferably the 2.6.39 variant rather than the 3.0
> > > variant, i.e. no magical -20°C offset, unless someone can explain where
> > > it comes from.
> > >
> > Sounds good.
> > 
> > Durga, do you have time to make those changes ?
> > 
> 
> I hope I can find some time in a couple of days.
> So I Will submit a patch for this.
> 
Never mind - I needed distraction and coded it in the last hour or so.

I'll test it tomorrow and send a patch unless I hit a major snag.

Thanks,
Guenter

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors