| Message ID | 20240820062010.2000873-1-raag.jadav@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v6] drm/i915/hwmon: expose fan speed | expand |
On Tue, Aug 20, 2024 at 11:50:10AM +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 > v2: Handle overflow, add mutex protection and ABI documentation > Aesthetic adjustments (Riana) > v3: Change rotations data type, ABI date and version > v4: Fix wakeref leak > Drop switch case and simplify hwm_fan_xx() (Andi) > v5: Rework time calculation, aesthetic adjustments (Andy) > v6: Drop overflow logic (Andy) > Aesthetic adjustments (Badal) But it still has an issue with 64-bit division on 32-bit platforms, right? ... > + /* > + * Calculate fan speed in RPM by time averaging two subsequent > + * readings in minutes. > + * RPM = number of rotations * msecs per minute / time in msecs > + */ > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); ...somewhere here?
On Tue, Aug 20, 2024 at 01:23:35PM +0300, Andy Shevchenko wrote: > On Tue, Aug 20, 2024 at 11:50:10AM +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 > > > v2: Handle overflow, add mutex protection and ABI documentation > > Aesthetic adjustments (Riana) > > v3: Change rotations data type, ABI date and version > > v4: Fix wakeref leak > > Drop switch case and simplify hwm_fan_xx() (Andi) > > v5: Rework time calculation, aesthetic adjustments (Andy) > > v6: Drop overflow logic (Andy) > > Aesthetic adjustments (Badal) > > But it still has an issue with 64-bit division on 32-bit platforms, right? > > ... > > > + /* > > + * Calculate fan speed in RPM by time averaging two subsequent > > + * readings in minutes. > > + * RPM = number of rotations * msecs per minute / time in msecs > > + */ > > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); > > ...somewhere here? Use div64_u64() instead? Or is there a better alternative? Raag
On Tue, Aug 20, 2024 at 04:12:46PM +0300, Raag Jadav wrote: > On Tue, Aug 20, 2024 at 01:23:35PM +0300, Andy Shevchenko wrote: > > On Tue, Aug 20, 2024 at 11:50:10AM +0530, Raag Jadav wrote: ... > > > v6: Drop overflow logic (Andy) > > > Aesthetic adjustments (Badal) > > > > But it still has an issue with 64-bit division on 32-bit platforms, right? ... > > > + /* > > > + * Calculate fan speed in RPM by time averaging two subsequent > > > + * readings in minutes. > > > + * RPM = number of rotations * msecs per minute / time in msecs > > > + */ > > > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); > > > > ...somewhere here? > > Use div64_u64() instead? Or is there a better alternative? I don't know the code well to offer anything better. Probably for a starter you can switch to use DIV_ROUND_UP_ULL().
Hi Raag, > + /* > + * Calculate fan speed in RPM by time averaging two subsequent > + * readings in minutes. > + * RPM = number of rotations * msecs per minute / time in msecs > + */ > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); once you find the correct operator here, please feel free to add: Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Normally these get me confused, I'm happy Andy is looking after it :-) Thanks, Andi
On Tue, Aug 20, 2024 at 05:06:39PM +0300, Andy Shevchenko wrote: > On Tue, Aug 20, 2024 at 04:12:46PM +0300, Raag Jadav wrote: > > On Tue, Aug 20, 2024 at 01:23:35PM +0300, Andy Shevchenko wrote: > > > On Tue, Aug 20, 2024 at 11:50:10AM +0530, Raag Jadav wrote: > > > > > > But it still has an issue with 64-bit division on 32-bit platforms, right? > > > > > > > + /* > > > > + * Calculate fan speed in RPM by time averaging two subsequent > > > > + * readings in minutes. > > > > + * RPM = number of rotations * msecs per minute / time in msecs > > > > + */ > > > > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); > > > > > > ...somewhere here? > > > > Use div64_u64() instead? Or is there a better alternative? > > I don't know the code well to offer anything better. Probably for a starter > you can switch to use DIV_ROUND_UP_ULL(). This one seems to work. Does it warrant an explanation here? Raag
On Thu, Aug 22, 2024 at 10:28:38AM +0300, Raag Jadav wrote: > On Tue, Aug 20, 2024 at 05:06:39PM +0300, Andy Shevchenko wrote: > > On Tue, Aug 20, 2024 at 04:12:46PM +0300, Raag Jadav wrote: > > > On Tue, Aug 20, 2024 at 01:23:35PM +0300, Andy Shevchenko wrote: > > > > On Tue, Aug 20, 2024 at 11:50:10AM +0530, Raag Jadav wrote: > > > > > > > > But it still has an issue with 64-bit division on 32-bit platforms, right? > > > > > > > > > + /* > > > > > + * Calculate fan speed in RPM by time averaging two subsequent > > > > > + * readings in minutes. > > > > > + * RPM = number of rotations * msecs per minute / time in msecs > > > > > + */ > > > > > + *val = DIV_ROUND_UP(rotations * (MSEC_PER_SEC * 60), time); > > > > > > > > ...somewhere here? > > > > > > Use div64_u64() instead? Or is there a better alternative? > > > > I don't know the code well to offer anything better. Probably for a starter > > you can switch to use DIV_ROUND_UP_ULL(). > > This one seems to work. Does it warrant an explanation here? You mean _ULL part? No.
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-rc5 next-20240823]
[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/20240820-135849
base: https://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git hwmon-next
patch link: https://lore.kernel.org/r/20240820062010.2000873-1-raag.jadav%40intel.com
patch subject: [PATCH v6] drm/i915/hwmon: expose fan speed
config: i386-allyesconfig (https://download.01.org/0day-ci/archive/20240826/202408260446.0EeOPMO6-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/20240826/202408260446.0EeOPMO6-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/202408260446.0EeOPMO6-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+0xe36): undefined reference to `__udivdi3'
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..0608bd4768e7 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 +622,69 @@ 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; + + if (attr == hwmon_fan_input && i915_mmio_reg_valid(hwmon->rg.fan_speed)) + return 0444; + + return 0; +} + +static int +hwm_fan_input_read(struct hwm_drvdata *ddat, 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; + int ret = 0; + + 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(); + + /* + * HW register value is accumulated count of pulses from + * PWM fan with the scale of 2 pulses per rotation. + */ + rotations = (reg_val - fi->reg_val_prev) / 2; + + time = jiffies_delta_to_msecs(time_now - fi->time_prev); + if (unlikely(!time)) { + ret = -EAGAIN; + goto exit; + } + + /* + * Calculate fan speed in RPM by time averaging two subsequent + * readings in minutes. + * 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 int +hwm_fan_read(struct hwm_drvdata *ddat, u32 attr, long *val) +{ + if (attr == hwmon_fan_input) + return hwm_fan_input_read(ddat, val); + + return -EOPNOTSUPP; +} + static umode_t hwm_is_visible(const void *drvdata, enum hwmon_sensor_types type, u32 attr, int channel) @@ -628,6 +700,8 @@ 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 +722,8 @@ 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 +815,14 @@ 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 +833,16 @@ 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);
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: Change rotations data type, ABI date and version v4: Fix wakeref leak Drop switch case and simplify hwm_fan_xx() (Andi) v5: Rework time calculation, aesthetic adjustments (Andy) v6: Drop overflow logic (Andy) Aesthetic adjustments (Badal) 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 | 88 +++++++++++++++++++ 3 files changed, 98 insertions(+)