Re: [PATCH v7 4/4] drm/xe/hwmon: Expose individual VRAM channel temperature

[Raag Jadav <raag.jadav@intel.com>]

On Tue, Jan 13, 2026 at 03:28:51PM +0530, Poosa, Karthik wrote:
> On 13-01-2026 13:51, Raag Jadav wrote:
> > On Tue, Jan 13, 2026 at 02:05:21AM +0530, Karthik Poosa wrote:
> > > Expose individual VRAM temperature attributes.
> > > Update Xe hwmon documentation for this entry.
> > ...
> > 
> > > @@ -255,6 +264,8 @@ static struct xe_reg xe_hwmon_get_reg(struct xe_hwmon *hwmon, enum xe_hwmon_reg
> > >   				return BMG_PACKAGE_TEMPERATURE;
> > >   			else if (channel == CHANNEL_VRAM)
> > >   				return BMG_VRAM_TEMPERATURE;
> > > +			else if (in_range(channel, CHANNEL_VRAM_N, CHANNEL_VRAM_N_MAX))
> > I think this should be
> > 
> > 	if (in_range(channel, CHANNEL_VRAM_N, MAX_VRAM_CHANNELS))
> > 
> > or did I miss something?
> I’ve ran IGT hwmon test and lm-sensors and didn't see any discrepancies
> there, all available channels are seen.

Sure, but shouldn't we fix the buggy logic?

Raag

> > > +				return BMG_VRAM_TEMPERATURE_N(channel - CHANNEL_VRAM_N);
> > >   		} else if (xe->info.platform == XE_DG2) {
> > >   			if (channel == CHANNEL_PKG)
> > >   				return PCU_CR_PACKAGE_TEMPERATURE;
> > ...
> > 
> > > @@ -1353,6 +1415,8 @@ static int xe_hwmon_read_label(struct device *dev,
> > >   			       enum hwmon_sensor_types type,
> > >   			       u32 attr, int channel, const char **str)
> > >   {
> > > +	struct xe_hwmon *hwmon = dev_get_drvdata(dev);
> > > +
> > >   	switch (type) {
> > >   	case hwmon_temp:
> > >   		if (channel == CHANNEL_PKG)
> > > @@ -1363,6 +1427,8 @@ static int xe_hwmon_read_label(struct device *dev,
> > >   			*str = "mctrl";
> > >   		else if (channel == CHANNEL_PCIE)
> > >   			*str = "pcie";
> > > +		else if (in_range(channel, CHANNEL_VRAM_N, CHANNEL_VRAM_N_MAX))
> > Ditto.
> > 
> > Raag
> > 
> > > +			*str = hwmon->temp.vram_label[channel - CHANNEL_VRAM_N];
> > >   		return 0;
> > >   	case hwmon_power:
> > >   	case hwmon_energy:
> > > -- 
> > > 2.25.1
> > >