diff mbox series

[v5] drm/i915/hwmon: expose fan speed

Message ID 20240812081538.1457396-1-raag.jadav@intel.com (mailing list archive)
State New, archived
Headers show
Series [v5] drm/i915/hwmon: expose fan speed | expand

Commit Message

Raag Jadav Aug. 12, 2024, 8:15 a.m. UTC
Add hwmon support for fan1_input attribute, which will expose fan speed
in RPM. With this in place we can monitor fan speed using lm-sensors tool.

$ sensors
i915-pci-0300
Adapter: PCI adapter
in0:         653.00 mV
fan1:        3833 RPM
power1:           N/A  (max =  43.00 W)
energy1:      32.02 kJ

v2: Handle overflow, add mutex protection and ABI documentation
    Aesthetic adjustments (Riana)
v3: Declare rotations as "long", change ABI date and version
    Add commenter name in changelog (Riana)
v4: Fix wakeref leak
    Drop switch case and simplify hwm_fan_xx() (Andi)
v5: Rework time calculation, aesthetic adjustments (Andy)

Signed-off-by: Raag Jadav <raag.jadav@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
---
 .../ABI/testing/sysfs-driver-intel-i915-hwmon |  8 ++
 drivers/gpu/drm/i915/gt/intel_gt_regs.h       |  2 +
 drivers/gpu/drm/i915/i915_hwmon.c             | 85 +++++++++++++++++++
 3 files changed, 95 insertions(+)

Comments

Andy Shevchenko Aug. 12, 2024, 1:15 p.m. UTC | #1
On Mon, Aug 12, 2024 at 01:45:38PM +0530, Raag Jadav wrote:
> Add hwmon support for fan1_input attribute, which will expose fan speed
> in RPM. With this in place we can monitor fan speed using lm-sensors tool.
> 
> $ sensors
> i915-pci-0300
> Adapter: PCI adapter
> in0:         653.00 mV
> fan1:        3833 RPM
> power1:           N/A  (max =  43.00 W)
> energy1:      32.02 kJ

...

> +static int
> +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> +{
> +	struct i915_hwmon *hwmon = ddat->hwmon;
> +	struct hwm_fan_info *fi = &ddat->fi;
> +	u64 rotations, time_now, time;
> +	intel_wakeref_t wakeref;
> +	u32 reg_val, pulses;
> +	int ret = 0;
> +
> +	if (attr != hwmon_fan_input)
> +		return -EOPNOTSUPP;
> +
> +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> +	mutex_lock(&hwmon->hwmon_lock);
> +
> +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> +	time_now = get_jiffies_64();

> +	/* Handle HW register overflow */
> +	if (reg_val >= fi->reg_val_prev)
> +		pulses = reg_val - fi->reg_val_prev;
> +	else
> +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;

Isn't it the abs_diff() reimplementation?

> +	/*
> +	 * HW register value is accumulated count of pulses from
> +	 * PWM fan with the scale of 2 pulses per rotation.
> +	 */
> +	rotations = pulses / 2;
> +
> +	time = jiffies_delta_to_msecs(time_now - fi->time_prev);
> +	if (unlikely(!time)) {
> +		ret = -EAGAIN;
> +		goto exit;
> +	}
> +
> +	/*
> +	 * Convert to minutes for calculating RPM.
> +	 * RPM = number of rotations * msecs per minute / time in msecs
> +	 */
> +	*val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time);
> +
> +	fi->reg_val_prev = reg_val;
> +	fi->time_prev = time_now;
> +exit:
> +	mutex_unlock(&hwmon->hwmon_lock);
> +	intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
> +	return ret;
> +}
Andi Shyti Aug. 12, 2024, 1:50 p.m. UTC | #2
Hi Raag,

> +static int
> +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> +{
> +	struct i915_hwmon *hwmon = ddat->hwmon;
> +	struct hwm_fan_info *fi = &ddat->fi;
> +	u64 rotations, time_now, time;
> +	intel_wakeref_t wakeref;
> +	u32 reg_val, pulses;
> +	int ret = 0;
> +
> +	if (attr != hwmon_fan_input)
> +		return -EOPNOTSUPP;
> +
> +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> +	mutex_lock(&hwmon->hwmon_lock);
> +
> +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> +	time_now = get_jiffies_64();
> +
> +	/* Handle HW register overflow */
> +	if (reg_val >= fi->reg_val_prev)
> +		pulses = reg_val - fi->reg_val_prev;
> +	else
> +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
> +
> +	/*
> +	 * HW register value is accumulated count of pulses from
> +	 * PWM fan with the scale of 2 pulses per rotation.
> +	 */
> +	rotations = pulses / 2;
> +
> +	time = jiffies_delta_to_msecs(time_now - fi->time_prev);
> +	if (unlikely(!time)) {
> +		ret = -EAGAIN;
> +		goto exit;
> +	}

Can you please add a comment describing how you obtain the speed
calculation?

Basically at every read you store the values. Is it possible that
we don't have reads for a long time and the register resets more
than once?

Thanks,
Andi

> +	/*
> +	 * Convert to minutes for calculating RPM.
> +	 * RPM = number of rotations * msecs per minute / time in msecs
> +	 */
> +	*val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time);
> +
> +	fi->reg_val_prev = reg_val;
> +	fi->time_prev = time_now;
> +exit:
> +	mutex_unlock(&hwmon->hwmon_lock);
> +	intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
> +	return ret;
> +}
kernel test robot Aug. 12, 2024, 9:36 p.m. UTC | #3
Hi Raag,

kernel test robot noticed the following build errors:

[auto build test ERROR on groeck-staging/hwmon-next]
[also build test ERROR on linus/master v6.11-rc3 next-20240812]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Raag-Jadav/drm-i915-hwmon-expose-fan-speed/20240812-161645
base:   https://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git hwmon-next
patch link:    https://lore.kernel.org/r/20240812081538.1457396-1-raag.jadav%40intel.com
patch subject: [PATCH v5] drm/i915/hwmon: expose fan speed
config: i386-randconfig-012-20240813 (https://download.01.org/0day-ci/archive/20240813/202408130500.SgCVoR2D-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240813/202408130500.SgCVoR2D-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408130500.SgCVoR2D-lkp@intel.com/

All errors (new ones prefixed by >>):

   ld: drivers/gpu/drm/i915/i915_hwmon.o: in function `hwm_fan_read':
>> drivers/gpu/drm/i915/i915_hwmon.c:675: undefined reference to `__udivdi3'


vim +675 drivers/gpu/drm/i915/i915_hwmon.c

   633	
   634	static int
   635	hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
   636	{
   637		struct i915_hwmon *hwmon = ddat->hwmon;
   638		struct hwm_fan_info *fi = &ddat->fi;
   639		u64 rotations, time_now, time;
   640		intel_wakeref_t wakeref;
   641		u32 reg_val, pulses;
   642		int ret = 0;
   643	
   644		if (attr != hwmon_fan_input)
   645			return -EOPNOTSUPP;
   646	
   647		wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
   648		mutex_lock(&hwmon->hwmon_lock);
   649	
   650		reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
   651		time_now = get_jiffies_64();
   652	
   653		/* Handle HW register overflow */
   654		if (reg_val >= fi->reg_val_prev)
   655			pulses = reg_val - fi->reg_val_prev;
   656		else
   657			pulses = UINT_MAX - fi->reg_val_prev + reg_val;
   658	
   659		/*
   660		 * HW register value is accumulated count of pulses from
   661		 * PWM fan with the scale of 2 pulses per rotation.
   662		 */
   663		rotations = pulses / 2;
   664	
   665		time = jiffies_delta_to_msecs(time_now - fi->time_prev);
   666		if (unlikely(!time)) {
   667			ret = -EAGAIN;
   668			goto exit;
   669		}
   670	
   671		/*
   672		 * Convert to minutes for calculating RPM.
   673		 * RPM = number of rotations * msecs per minute / time in msecs
   674		 */
 > 675		*val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time);
   676	
   677		fi->reg_val_prev = reg_val;
   678		fi->time_prev = time_now;
   679	exit:
   680		mutex_unlock(&hwmon->hwmon_lock);
   681		intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
   682		return ret;
   683	}
   684
kernel test robot Aug. 13, 2024, 12:38 a.m. UTC | #4
Hi Raag,

kernel test robot noticed the following build errors:

[auto build test ERROR on groeck-staging/hwmon-next]
[also build test ERROR on linus/master v6.11-rc3 next-20240812]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Raag-Jadav/drm-i915-hwmon-expose-fan-speed/20240812-161645
base:   https://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git hwmon-next
patch link:    https://lore.kernel.org/r/20240812081538.1457396-1-raag.jadav%40intel.com
patch subject: [PATCH v5] drm/i915/hwmon: expose fan speed
config: i386-allyesconfig (https://download.01.org/0day-ci/archive/20240813/202408130800.XtY6XxQ5-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240813/202408130800.XtY6XxQ5-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202408130800.XtY6XxQ5-lkp@intel.com/

All errors (new ones prefixed by >>):

   ld: drivers/gpu/drm/i915/i915_hwmon.o: in function `hwm_read':
>> i915_hwmon.c:(.text+0xe60): undefined reference to `__udivdi3'
Raag Jadav Aug. 13, 2024, 5:45 a.m. UTC | #5
On Mon, Aug 12, 2024 at 04:15:14PM +0300, Andy Shevchenko wrote:
> On Mon, Aug 12, 2024 at 01:45:38PM +0530, Raag Jadav wrote:
> > Add hwmon support for fan1_input attribute, which will expose fan speed
> > in RPM. With this in place we can monitor fan speed using lm-sensors tool.
> > 
> > $ sensors
> > i915-pci-0300
> > Adapter: PCI adapter
> > in0:         653.00 mV
> > fan1:        3833 RPM
> > power1:           N/A  (max =  43.00 W)
> > energy1:      32.02 kJ
> 
> ...
> 
> > +static int
> > +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> > +{
> > +	struct i915_hwmon *hwmon = ddat->hwmon;
> > +	struct hwm_fan_info *fi = &ddat->fi;
> > +	u64 rotations, time_now, time;
> > +	intel_wakeref_t wakeref;
> > +	u32 reg_val, pulses;
> > +	int ret = 0;
> > +
> > +	if (attr != hwmon_fan_input)
> > +		return -EOPNOTSUPP;
> > +
> > +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> > +	mutex_lock(&hwmon->hwmon_lock);
> > +
> > +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> > +	time_now = get_jiffies_64();
> 
> > +	/* Handle HW register overflow */
> > +	if (reg_val >= fi->reg_val_prev)
> > +		pulses = reg_val - fi->reg_val_prev;
> > +	else
> > +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
> 
> Isn't it the abs_diff() reimplementation?

Not exactly. This is specific to 32 bit register overflow, so we count
from max value.

Raag
Raag Jadav Aug. 13, 2024, 6:30 a.m. UTC | #6
On Mon, Aug 12, 2024 at 03:50:47PM +0200, Andi Shyti wrote:
> Hi Raag,
> 
> > +static int
> > +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> > +{
> > +	struct i915_hwmon *hwmon = ddat->hwmon;
> > +	struct hwm_fan_info *fi = &ddat->fi;
> > +	u64 rotations, time_now, time;
> > +	intel_wakeref_t wakeref;
> > +	u32 reg_val, pulses;
> > +	int ret = 0;
> > +
> > +	if (attr != hwmon_fan_input)
> > +		return -EOPNOTSUPP;
> > +
> > +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> > +	mutex_lock(&hwmon->hwmon_lock);
> > +
> > +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> > +	time_now = get_jiffies_64();
> > +
> > +	/* Handle HW register overflow */
> > +	if (reg_val >= fi->reg_val_prev)
> > +		pulses = reg_val - fi->reg_val_prev;
> > +	else
> > +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
> > +
> > +	/*
> > +	 * HW register value is accumulated count of pulses from
> > +	 * PWM fan with the scale of 2 pulses per rotation.
> > +	 */
> > +	rotations = pulses / 2;
> > +
> > +	time = jiffies_delta_to_msecs(time_now - fi->time_prev);
> > +	if (unlikely(!time)) {
> > +		ret = -EAGAIN;
> > +		goto exit;
> > +	}
> 
> Can you please add a comment describing how you obtain the speed
> calculation?

That's what I initially tried but ended up dropping it in favour of RPM
formula below, which I found to be doing a better job of explaining than
a few lines of description.

> Basically at every read you store the values. Is it possible that
> we don't have reads for a long time and the register resets more
> than once?

Considering a fan continuously running at higher speeds (for example 4000 RPM
which is quite optimistic), with the scale of 2 pulses per rotation, a 32 bit
register will take around a year to overflow, which is more than most usecases
I could think of.

Raag

> > +	/*
> > +	 * Convert to minutes for calculating RPM.
> > +	 * RPM = number of rotations * msecs per minute / time in msecs
> > +	 */
> > +	*val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time);
> > +
> > +	fi->reg_val_prev = reg_val;
> > +	fi->time_prev = time_now;
> > +exit:
> > +	mutex_unlock(&hwmon->hwmon_lock);
> > +	intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
> > +	return ret;
> > +}
Andy Shevchenko Aug. 13, 2024, 8:27 a.m. UTC | #7
On Tue, Aug 13, 2024 at 08:45:19AM +0300, Raag Jadav wrote:
> On Mon, Aug 12, 2024 at 04:15:14PM +0300, Andy Shevchenko wrote:
> > On Mon, Aug 12, 2024 at 01:45:38PM +0530, Raag Jadav wrote:

...

> > > +static int
> > > +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> > > +{
> > > +	struct i915_hwmon *hwmon = ddat->hwmon;
> > > +	struct hwm_fan_info *fi = &ddat->fi;
> > > +	u64 rotations, time_now, time;
> > > +	intel_wakeref_t wakeref;
> > > +	u32 reg_val, pulses;
> > > +	int ret = 0;
> > > +
> > > +	if (attr != hwmon_fan_input)
> > > +		return -EOPNOTSUPP;
> > > +
> > > +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> > > +	mutex_lock(&hwmon->hwmon_lock);
> > > +
> > > +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> > > +	time_now = get_jiffies_64();
> > 
> > > +	/* Handle HW register overflow */
> > > +	if (reg_val >= fi->reg_val_prev)
> > > +		pulses = reg_val - fi->reg_val_prev;
> > > +	else
> > > +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
> > 
> > Isn't it the abs_diff() reimplementation?
> 
> Not exactly. This is specific to 32 bit register overflow, so we count
> from max value.

I see. But since you have the both variables of u32, why:
1) UINT_MAX?
2) Not simply using

	pulses = reg_val - fi->reg_val_prev;

which will wrap over correctly?

Note, in your case (in comparison to the wrap over variant) the off-by-one is
present. Is it on purpose?
Raag Jadav Aug. 13, 2024, 10:03 a.m. UTC | #8
On Tue, Aug 13, 2024 at 11:27:22AM +0300, Andy Shevchenko wrote:
> On Tue, Aug 13, 2024 at 08:45:19AM +0300, Raag Jadav wrote:
> > On Mon, Aug 12, 2024 at 04:15:14PM +0300, Andy Shevchenko wrote:
> > > On Mon, Aug 12, 2024 at 01:45:38PM +0530, Raag Jadav wrote:
> 
> ...
> 
> > > > +static int
> > > > +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
> > > > +{
> > > > +	struct i915_hwmon *hwmon = ddat->hwmon;
> > > > +	struct hwm_fan_info *fi = &ddat->fi;
> > > > +	u64 rotations, time_now, time;
> > > > +	intel_wakeref_t wakeref;
> > > > +	u32 reg_val, pulses;
> > > > +	int ret = 0;
> > > > +
> > > > +	if (attr != hwmon_fan_input)
> > > > +		return -EOPNOTSUPP;
> > > > +
> > > > +	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
> > > > +	mutex_lock(&hwmon->hwmon_lock);
> > > > +
> > > > +	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
> > > > +	time_now = get_jiffies_64();
> > > 
> > > > +	/* Handle HW register overflow */
> > > > +	if (reg_val >= fi->reg_val_prev)
> > > > +		pulses = reg_val - fi->reg_val_prev;
> > > > +	else
> > > > +		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
> > > 
> > > Isn't it the abs_diff() reimplementation?
> > 
> > Not exactly. This is specific to 32 bit register overflow, so we count
> > from max value.
> 
> I see. But since you have the both variables of u32, why:
> 1) UINT_MAX?
> 2) Not simply using
> 
> 	pulses = reg_val - fi->reg_val_prev;
> 
> which will wrap over correctly?

Agree. Will update.

Raag
Andi Shyti Aug. 21, 2024, 2:58 p.m. UTC | #9
> > > +	/*
> > > +	 * HW register value is accumulated count of pulses from
> > > +	 * PWM fan with the scale of 2 pulses per rotation.
> > > +	 */
> > > +	rotations = pulses / 2;
> > > +
> > > +	time = jiffies_delta_to_msecs(time_now - fi->time_prev);
> > > +	if (unlikely(!time)) {
> > > +		ret = -EAGAIN;
> > > +		goto exit;
> > > +	}
> > 
> > Can you please add a comment describing how you obtain the speed
> > calculation?
> 
> That's what I initially tried but ended up dropping it in favour of RPM
> formula below, which I found to be doing a better job of explaining than
> a few lines of description.
> 
> > Basically at every read you store the values. Is it possible that
> > we don't have reads for a long time and the register resets more
> > than once?
> 
> Considering a fan continuously running at higher speeds (for example 4000 RPM
> which is quite optimistic), with the scale of 2 pulses per rotation, a 32 bit
> register will take around a year to overflow, which is more than most usecases
> I could think of.

Which can be considered as a worse case scenario. I would have
preferred here a runtime calculation, which means read now, wait
a bit, read again and calculate. The read might be slow, but
efficient.

Anyway, your argument makes sense, so that I'm not going to push
on this, I already r-b'ed it.

Thanks,
Andi
diff mbox series

Patch

diff --git a/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon b/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon
index 92fe7c5c5ac1..be4141a7522f 100644
--- a/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon
+++ b/Documentation/ABI/testing/sysfs-driver-intel-i915-hwmon
@@ -75,3 +75,11 @@  Description:	RO. Energy input of device or gt in microjoules.
 		for the gt.
 
 		Only supported for particular Intel i915 graphics platforms.
+
+What:		/sys/bus/pci/drivers/i915/.../hwmon/hwmon<i>/fan1_input
+Date:		November 2024
+KernelVersion:	6.12
+Contact:	intel-gfx@lists.freedesktop.org
+Description:	RO. Fan speed of device in RPM.
+
+		Only supported for particular Intel i915 graphics platforms.
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_regs.h b/drivers/gpu/drm/i915/gt/intel_gt_regs.h
index e42b3a5d4e63..57a3c83d3655 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_regs.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_regs.h
@@ -1553,6 +1553,8 @@ 
 #define VLV_RENDER_C0_COUNT			_MMIO(0x138118)
 #define VLV_MEDIA_C0_COUNT			_MMIO(0x13811c)
 
+#define PCU_PWM_FAN_SPEED			_MMIO(0x138140)
+
 #define GEN12_RPSTAT1				_MMIO(0x1381b4)
 #define   GEN12_VOLTAGE_MASK			REG_GENMASK(10, 0)
 #define   GEN12_CAGF_MASK			REG_GENMASK(19, 11)
diff --git a/drivers/gpu/drm/i915/i915_hwmon.c b/drivers/gpu/drm/i915/i915_hwmon.c
index 49db3e09826c..a867b81a0900 100644
--- a/drivers/gpu/drm/i915/i915_hwmon.c
+++ b/drivers/gpu/drm/i915/i915_hwmon.c
@@ -5,6 +5,7 @@ 
 
 #include <linux/hwmon.h>
 #include <linux/hwmon-sysfs.h>
+#include <linux/jiffies.h>
 #include <linux/types.h>
 
 #include "i915_drv.h"
@@ -36,6 +37,7 @@  struct hwm_reg {
 	i915_reg_t pkg_rapl_limit;
 	i915_reg_t energy_status_all;
 	i915_reg_t energy_status_tile;
+	i915_reg_t fan_speed;
 };
 
 struct hwm_energy_info {
@@ -43,11 +45,17 @@  struct hwm_energy_info {
 	long accum_energy;			/* Accumulated energy for energy1_input */
 };
 
+struct hwm_fan_info {
+	u32 reg_val_prev;
+	u64 time_prev;
+};
+
 struct hwm_drvdata {
 	struct i915_hwmon *hwmon;
 	struct intel_uncore *uncore;
 	struct device *hwmon_dev;
 	struct hwm_energy_info ei;		/*  Energy info for energy1_input */
+	struct hwm_fan_info fi;			/*  Fan info for fan1_input */
 	char name[12];
 	int gt_n;
 	bool reset_in_progress;
@@ -276,6 +284,7 @@  static const struct hwmon_channel_info * const hwm_info[] = {
 	HWMON_CHANNEL_INFO(power, HWMON_P_MAX | HWMON_P_RATED_MAX | HWMON_P_CRIT),
 	HWMON_CHANNEL_INFO(energy, HWMON_E_INPUT),
 	HWMON_CHANNEL_INFO(curr, HWMON_C_CRIT),
+	HWMON_CHANNEL_INFO(fan, HWMON_F_INPUT),
 	NULL
 };
 
@@ -613,6 +622,66 @@  hwm_curr_write(struct hwm_drvdata *ddat, u32 attr, long val)
 	}
 }
 
+static umode_t
+hwm_fan_is_visible(const struct hwm_drvdata *ddat, u32 attr)
+{
+	struct i915_hwmon *hwmon = ddat->hwmon;
+
+	return attr == hwmon_fan_input &&
+	       i915_mmio_reg_valid(hwmon->rg.fan_speed) ? 0444 : 0;
+}
+
+static int
+hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val)
+{
+	struct i915_hwmon *hwmon = ddat->hwmon;
+	struct hwm_fan_info *fi = &ddat->fi;
+	u64 rotations, time_now, time;
+	intel_wakeref_t wakeref;
+	u32 reg_val, pulses;
+	int ret = 0;
+
+	if (attr != hwmon_fan_input)
+		return -EOPNOTSUPP;
+
+	wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
+	mutex_lock(&hwmon->hwmon_lock);
+
+	reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.fan_speed);
+	time_now = get_jiffies_64();
+
+	/* Handle HW register overflow */
+	if (reg_val >= fi->reg_val_prev)
+		pulses = reg_val - fi->reg_val_prev;
+	else
+		pulses = UINT_MAX - fi->reg_val_prev + reg_val;
+
+	/*
+	 * HW register value is accumulated count of pulses from
+	 * PWM fan with the scale of 2 pulses per rotation.
+	 */
+	rotations = pulses / 2;
+
+	time = jiffies_delta_to_msecs(time_now - fi->time_prev);
+	if (unlikely(!time)) {
+		ret = -EAGAIN;
+		goto exit;
+	}
+
+	/*
+	 * Convert to minutes for calculating RPM.
+	 * RPM = number of rotations * msecs per minute / time in msecs
+	 */
+	*val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time);
+
+	fi->reg_val_prev = reg_val;
+	fi->time_prev = time_now;
+exit:
+	mutex_unlock(&hwmon->hwmon_lock);
+	intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
+	return ret;
+}
+
 static umode_t
 hwm_is_visible(const void *drvdata, enum hwmon_sensor_types type,
 	       u32 attr, int channel)
@@ -628,6 +697,8 @@  hwm_is_visible(const void *drvdata, enum hwmon_sensor_types type,
 		return hwm_energy_is_visible(ddat, attr);
 	case hwmon_curr:
 		return hwm_curr_is_visible(ddat, attr);
+	case hwmon_fan:
+		return hwm_fan_is_visible(ddat, attr);
 	default:
 		return 0;
 	}
@@ -648,6 +719,8 @@  hwm_read(struct device *dev, enum hwmon_sensor_types type, u32 attr,
 		return hwm_energy_read(ddat, attr, val);
 	case hwmon_curr:
 		return hwm_curr_read(ddat, attr, val);
+	case hwmon_fan:
+		return hwm_fan_read(ddat, attr, val);
 	default:
 		return -EOPNOTSUPP;
 	}
@@ -739,12 +812,14 @@  hwm_get_preregistration_info(struct drm_i915_private *i915)
 		hwmon->rg.pkg_rapl_limit = PCU_PACKAGE_RAPL_LIMIT;
 		hwmon->rg.energy_status_all = PCU_PACKAGE_ENERGY_STATUS;
 		hwmon->rg.energy_status_tile = INVALID_MMIO_REG;
+		hwmon->rg.fan_speed = PCU_PWM_FAN_SPEED;
 	} else {
 		hwmon->rg.pkg_power_sku_unit = INVALID_MMIO_REG;
 		hwmon->rg.pkg_power_sku = INVALID_MMIO_REG;
 		hwmon->rg.pkg_rapl_limit = INVALID_MMIO_REG;
 		hwmon->rg.energy_status_all = INVALID_MMIO_REG;
 		hwmon->rg.energy_status_tile = INVALID_MMIO_REG;
+		hwmon->rg.fan_speed = INVALID_MMIO_REG;
 	}
 
 	with_intel_runtime_pm(uncore->rpm, wakeref) {
@@ -755,6 +830,16 @@  hwm_get_preregistration_info(struct drm_i915_private *i915)
 		if (i915_mmio_reg_valid(hwmon->rg.pkg_power_sku_unit))
 			val_sku_unit = intel_uncore_read(uncore,
 							 hwmon->rg.pkg_power_sku_unit);
+
+		/*
+		 * Store the initial fan register value, so that we can use it for
+		 * initial fan speed calculation.
+		 */
+		if (i915_mmio_reg_valid(hwmon->rg.fan_speed)) {
+			ddat->fi.reg_val_prev = intel_uncore_read(uncore,
+								  hwmon->rg.fan_speed);
+			ddat->fi.time_prev = get_jiffies_64();
+		}
 	}
 
 	hwmon->scl_shift_power = REG_FIELD_GET(PKG_PWR_UNIT, val_sku_unit);