[lm-sensors] [PATCH v3] hwmon: (it87) Make temp3 attribute conditional for IT8782F

[Guenter Roeck]

On Fri, 2012-05-11 at 17:07 -0400, Jean Delvare wrote:
> Hi Guenter,
> 
> On Fri, 11 May 2012 08:49:43 -0700, Guenter Roeck wrote:
> > From: Guenter Roeck <linux at roeck-us.net>
> > 
> > On IT8782F, temp3 is only supported if UART6 is disabled.
> > 
> > Signed-off-by: Guenter Roeck <linux at roeck-us.net>
> > ---
> > v3: Resolve conflicts with previous patches.
> >     Only remove temp sysfs attributes if the to-be-removed sensor exists,
> >     to be in sync with the other attribute removal code.
> > v2: Drop condition for IT8783F. On that chip, UART6 may be routed via JP4 to
> >     a different set of pins, so we can not reliably detect if temp3 and RI6
> >     overlap or not. 
> > 
> >  drivers/hwmon/it87.c |   79 +++++++++++++++++++++++++++++++++++++------------
> >  1 files changed, 59 insertions(+), 20 deletions(-)
> > 
> > diff --git a/drivers/hwmon/it87.c b/drivers/hwmon/it87.c
> > index 88771a5..1385320 100644
> > --- a/drivers/hwmon/it87.c
> > +++ b/drivers/hwmon/it87.c
> > (...)
> > @@ -2016,6 +2040,11 @@ static int __devinit it87_probe(struct platform_device *pdev)
> >  			data->in_scaled |= (1 << 7);	/* in7 is VCCH5V */
> >  	}
> >  
> > +	data->has_temp = 0x07;
> > +	if (sio_data->skip_temp &&
> > +	    !(it87_read_value(data, IT87_REG_TEMP_EXTRA) & 0x80))
> > +		data->has_temp &= ~sio_data->skip_temp;
> 
> I don't like this. Right now it works because temp3 is the only input
> that can be disabled. If there ever is another then it will break and
> we may not notice. IMHO it should look more like:
> 
> 	if ((sio_data->skip_temp & (1 << 2)) &&
> 	    !(it87_read_value(data, IT87_REG_TEMP_EXTRA) & 0x80))
> 		data->has_temp &= ~(1 << 2);
> 
> I would also suggest an additional test on chip type. Right now the
> IT8782F is the only one setting sio_data->skip_temp but that might
> change in the future.
> 
Agreed. I'll send out a revised version in a minute.

Thanks,
Guenter