[PATCH v5 4/4] drm/xe/hwmon: Expose individual vram channel temperature

[Karthik Poosa <karthik.poosa@intel.com>]

Expose individual VRAM temperature attributes.
Update Xe hwmon documentation for this entry.

v2:
 - Avoid using default switch case for VRAM individual temperatures.
 - Append labels with vram number.
 - Update kernel version in Xe hwmon documentation.

v3:
 - Add missing brackets in Xe hwmon documentation from vram sysfs.
 - Reorder BMG_VRAM_TEMPERATURE_N macro in xe_pcode_regs.h.
 - Add api to check if vram is available on the channel.

v4:
 - Improve VRAM label handling to eliminate temp variable by
   introducing a dedicated array vram_label in xe_hwmon_thermal_info.
 - Remove a magic number.
 - Change the label from vram_X to vram_ch_X.

Signed-off-by: Karthik Poosa <karthik.poosa@intel.com>
---
 .../ABI/testing/sysfs-driver-intel-xe-hwmon   | 22 +++++++
 drivers/gpu/drm/xe/regs/xe_pcode_regs.h       |  3 +
 drivers/gpu/drm/xe/xe_hwmon.c                 | 65 ++++++++++++++++++-
 3 files changed, 89 insertions(+), 1 deletion(-)

diff --git a/Documentation/ABI/testing/sysfs-driver-intel-xe-hwmon b/Documentation/ABI/testing/sysfs-driver-intel-xe-hwmon
index 6e21bebf0e0d..55ab45f669ac 100644
--- a/Documentation/ABI/testing/sysfs-driver-intel-xe-hwmon
+++ b/Documentation/ABI/testing/sysfs-driver-intel-xe-hwmon
@@ -211,6 +211,28 @@ KernelVersion:	7.0
 Contact:	intel-xe@lists.freedesktop.org
 Description:	RO. GPU PCIe temperature in millidegree Celsius.
 
+What:		/sys/bus/pci/drivers/xe/.../hwmon/hwmon<i>/temp[6-21]_crit
+Date:		January 2026
+KernelVersion:	7.0
+Contact:	intel-xe@lists.freedesktop.org
+Description:	RO. VRAM channel critical temperature in millidegree Celsius.
+
+		Only supported for particular Intel Xe graphics platforms.
+
+What:		/sys/bus/pci/drivers/xe/.../hwmon/hwmon<i>/temp[6-21]_emergency
+Date:		January 2026
+KernelVersion:	7.0
+Contact:	intel-xe@lists.freedesktop.org
+Description:	RO. VRAM channel shutdown temperature in millidegree Celsius.
+
+		Only supported for particular Intel Xe graphics platforms.
+
+What:		/sys/bus/pci/drivers/xe/.../hwmon/hwmon<i>/temp[6-21]_input
+Date:		January 2026
+KernelVersion:	7.0
+Contact:	intel-xe@lists.freedesktop.org
+Description:	RO. VRAM channel temperature in millidegree Celsius.
+
 		Only supported for particular Intel Xe graphics platforms.
 
 What:		/sys/bus/pci/drivers/xe/.../hwmon/hwmon<i>/fan1_input
diff --git a/drivers/gpu/drm/xe/regs/xe_pcode_regs.h b/drivers/gpu/drm/xe/regs/xe_pcode_regs.h
index fb097607b86c..6627591c05b7 100644
--- a/drivers/gpu/drm/xe/regs/xe_pcode_regs.h
+++ b/drivers/gpu/drm/xe/regs/xe_pcode_regs.h
@@ -22,6 +22,9 @@
 #define BMG_FAN_2_SPEED				XE_REG(0x138170)
 #define BMG_FAN_3_SPEED				XE_REG(0x1381a0)
 #define BMG_VRAM_TEMPERATURE			XE_REG(0x1382c0)
+#define BMG_VRAM_TEMPERATURE_N(n)		XE_REG(0x138260 + (n))
+#define   TEMP_MASK_VRAM_N			REG_GENMASK(30, 8)
+#define   TEMP_SIGN_MASK			BIT(31)
 #define BMG_PACKAGE_TEMPERATURE			XE_REG(0x138434)
 
 #endif /* _XE_PCODE_REGS_H_ */
diff --git a/drivers/gpu/drm/xe/xe_hwmon.c b/drivers/gpu/drm/xe/xe_hwmon.c
index 2bb67471b755..2403687a73fe 100644
--- a/drivers/gpu/drm/xe/xe_hwmon.c
+++ b/drivers/gpu/drm/xe/xe_hwmon.c
@@ -39,12 +39,16 @@ enum xe_hwmon_reg_operation {
 	REG_READ64,
 };
 
+#define MAX_VRAM_CHANNELS      (16)
+
 enum xe_hwmon_channel {
 	CHANNEL_CARD,
 	CHANNEL_PKG,
 	CHANNEL_VRAM,
 	CHANNEL_MCTRL,
 	CHANNEL_PCIE,
+	CHANNEL_VRAM_N,
+	CHANNEL_VRAM_N_MAX = CHANNEL_VRAM_N + MAX_VRAM_CHANNELS,
 	CHANNEL_MAX,
 };
 
@@ -106,6 +110,7 @@ enum sensor_attr_power {
 #define TEMP_INDEX_MCTRL	2
 #define PCIE_SENSOR_GROUP_ID	0x2
 #define PCIE_SENSOR_MASK	REG_GENMASK(31, 16)
+#define MAX_LABEL_SIZE		16
 
 /**
  * struct xe_hwmon_energy_info - to accumulate energy
@@ -141,6 +146,8 @@ struct xe_hwmon_thermal_info {
 	u8 count;
 	/** @value: signed value from each sensor */
 	s8 value[U8_MAX];
+	/** @vram_label: vram label names */
+	char vram_label[MAX_VRAM_CHANNELS][MAX_LABEL_SIZE];
 };
 
 /**
@@ -257,6 +264,9 @@ 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 (channel >= CHANNEL_VRAM_N && channel <= CHANNEL_VRAM_N_MAX)
+				return BMG_VRAM_TEMPERATURE_N(((channel - CHANNEL_VRAM_N) *
+								sizeof(u32)));
 		} else if (xe->info.platform == XE_DG2) {
 			if (channel == CHANNEL_PKG)
 				return PCU_CR_PACKAGE_TEMPERATURE;
@@ -716,6 +726,22 @@ static const struct hwmon_channel_info * const hwmon_info[] = {
 			   HWMON_T_MAX,
 			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
 			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
+			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL,
 			   HWMON_T_CRIT | HWMON_T_EMERGENCY | HWMON_T_INPUT | HWMON_T_LABEL),
 	HWMON_CHANNEL_INFO(power, HWMON_P_MAX | HWMON_P_RATED_MAX | HWMON_P_LABEL | HWMON_P_CRIT |
 			   HWMON_P_CAP,
@@ -890,6 +916,21 @@ static void xe_hwmon_get_voltage(struct xe_hwmon *hwmon, int channel, long *valu
 	*value = DIV_ROUND_CLOSEST(REG_FIELD_GET(VOLTAGE_MASK, reg_val) * 2500, SF_VOLTAGE);
 }
 
+static inline bool is_vram_ch_available(struct xe_hwmon *hwmon, int channel)
+{
+	struct xe_reg vram_ch_temp;
+	struct xe_mmio *mmio = xe_root_tile_mmio(hwmon->xe);
+
+	vram_ch_temp = xe_hwmon_get_reg(hwmon, REG_TEMP, channel);
+	if (xe_reg_is_valid(vram_ch_temp) && xe_mmio_read32(mmio, vram_ch_temp)) {
+		/* Create label only for available vram channel */
+		sprintf(hwmon->temp.vram_label[channel - CHANNEL_VRAM_N], "vram_ch_%d",
+			(channel - CHANNEL_VRAM_N));
+		return 1;
+	}
+	return 0;
+}
+
 static umode_t
 xe_hwmon_temp_is_visible(struct xe_hwmon *hwmon, u32 attr, int channel)
 {
@@ -903,6 +944,8 @@ xe_hwmon_temp_is_visible(struct xe_hwmon *hwmon, u32 attr, int channel)
 		case CHANNEL_MCTRL:
 		case CHANNEL_PCIE:
 			return hwmon->temp.count ? 0444 : 0;
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
+			return is_vram_ch_available(hwmon, channel) ? 0444 : 0;
 		default:
 			return 0;
 		}
@@ -915,6 +958,8 @@ xe_hwmon_temp_is_visible(struct xe_hwmon *hwmon, u32 attr, int channel)
 		case CHANNEL_MCTRL:
 		case CHANNEL_PCIE:
 			return hwmon->temp.count ? 0444 : 0;
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
+			return is_vram_ch_available(hwmon, channel) ? 0444 : 0;
 		default:
 			return 0;
 		}
@@ -935,6 +980,8 @@ xe_hwmon_temp_is_visible(struct xe_hwmon *hwmon, u32 attr, int channel)
 		case CHANNEL_MCTRL:
 		case CHANNEL_PCIE:
 			return hwmon->temp.count ? 0444 : 0;
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
+			return is_vram_ch_available(hwmon, channel) ? 0444 : 0;
 		default:
 			return 0;
 		}
@@ -963,6 +1010,16 @@ xe_hwmon_temp_read(struct xe_hwmon *hwmon, u32 attr, int channel, long *val)
 			return get_mc_temp(hwmon, val);
 		case CHANNEL_PCIE:
 			return get_pcie_temp(hwmon, val);
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
+			reg_val = xe_mmio_read32(mmio, xe_hwmon_get_reg(hwmon, REG_TEMP, channel));
+			/*
+			 * This temperature format is bit 31 for sign, bits [30:8] for whole number
+			 * and bits [7:0] for fraction
+			 */
+			*val = (s32)(REG_FIELD_GET(TEMP_MASK_VRAM_N, reg_val)) *
+				(REG_FIELD_GET(TEMP_SIGN_MASK, reg_val) ? -1 : 1) *
+				 MILLIDEGREE_PER_DEGREE;
+			return 0;
 		default:
 			return -EOPNOTSUPP;
 		}
@@ -974,6 +1031,7 @@ xe_hwmon_temp_read(struct xe_hwmon *hwmon, u32 attr, int channel, long *val)
 			*val = hwmon->temp.limit[TEMP_LIMIT_PKG_SHUTDOWN] * MILLIDEGREE_PER_DEGREE;
 			return 0;
 		case CHANNEL_VRAM:
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
 			*val = hwmon->temp.limit[TEMP_LIMIT_MEM_SHUTDOWN] * MILLIDEGREE_PER_DEGREE;
 			return 0;
 		default:
@@ -987,6 +1045,7 @@ xe_hwmon_temp_read(struct xe_hwmon *hwmon, u32 attr, int channel, long *val)
 			*val = hwmon->temp.limit[TEMP_LIMIT_PKG_CRIT] * MILLIDEGREE_PER_DEGREE;
 			return 0;
 		case CHANNEL_VRAM:
+		case CHANNEL_VRAM_N...CHANNEL_VRAM_N_MAX:
 			*val = hwmon->temp.limit[TEMP_LIMIT_MEM_CRIT] * MILLIDEGREE_PER_DEGREE;
 			return 0;
 		default:
@@ -1356,16 +1415,20 @@ 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)
 			*str = "pkg";
 		else if (channel == CHANNEL_VRAM)
-			*str = "vram";
+			*str = "vram_avg";
 		else if (channel == CHANNEL_MCTRL)
 			*str = "mctrl";
 		else if (channel == CHANNEL_PCIE)
 			*str = "pcie";
+		else if (channel >= CHANNEL_VRAM_N && channel <= CHANNEL_VRAM_N_MAX)
+			*str = hwmon->temp.vram_label[channel - CHANNEL_VRAM_N];
 		return 0;
 	case hwmon_power:
 	case hwmon_energy:
-- 
2.25.1