linux/drivers/gpu/drm/i915/i915_hwmon.c

995 lines
25 KiB
C
Raw Permalink Normal View History

// SPDX-License-Identifier: MIT
/*
* Copyright © 2022 Intel Corporation
*/
#include <linux/hwmon.h>
#include <linux/hwmon-sysfs.h>
#include <linux/jiffies.h>
#include <linux/types.h>
#include <linux/units.h>
#include "i915_drv.h"
#include "i915_hwmon.h"
#include "i915_reg.h"
#include "intel_mchbar_regs.h"
#include "intel_pcode.h"
#include "gt/intel_gt.h"
#include "gt/intel_gt_regs.h"
/*
* SF_* - scale factors for particular quantities according to hwmon spec.
* - voltage - millivolts
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
* - power - microwatts
* - curr - milliamperes
* - energy - microjoules
* - time - milliseconds
*/
#define SF_VOLTAGE 1000
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
#define SF_POWER 1000000
#define SF_CURR 1000
#define SF_ENERGY 1000000
#define SF_TIME 1000
struct hwm_reg {
i915_reg_t gt_perf_status;
i915_reg_t pkg_temp;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
i915_reg_t pkg_power_sku_unit;
i915_reg_t pkg_power_sku;
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 {
u32 reg_val_prev;
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;
wait_queue_head_t waitq;
};
struct i915_hwmon {
struct hwm_drvdata ddat;
struct hwm_drvdata ddat_gt[I915_MAX_GT];
struct mutex hwmon_lock; /* counter overflow logic and rmw */
struct hwm_reg rg;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
int scl_shift_power;
int scl_shift_energy;
int scl_shift_time;
};
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
static void
hwm_locked_with_pm_intel_uncore_rmw(struct hwm_drvdata *ddat,
i915_reg_t reg, u32 clear, u32 set)
{
struct i915_hwmon *hwmon = ddat->hwmon;
struct intel_uncore *uncore = ddat->uncore;
intel_wakeref_t wakeref;
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
with_intel_runtime_pm(uncore->rpm, wakeref) {
mutex_lock(&hwmon->hwmon_lock);
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
intel_uncore_rmw(uncore, reg, clear, set);
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
mutex_unlock(&hwmon->hwmon_lock);
}
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
}
/*
* This function's return type of u64 allows for the case where the scaling
* of the field taken from the 32-bit register value might cause a result to
* exceed 32 bits.
*/
static u64
hwm_field_read_and_scale(struct hwm_drvdata *ddat, i915_reg_t rgadr,
u32 field_msk, int nshift, u32 scale_factor)
{
struct intel_uncore *uncore = ddat->uncore;
intel_wakeref_t wakeref;
u32 reg_value;
with_intel_runtime_pm(uncore->rpm, wakeref)
reg_value = intel_uncore_read(uncore, rgadr);
reg_value = REG_FIELD_GET(field_msk, reg_value);
return mul_u64_u32_shr(reg_value, scale_factor, nshift);
}
/*
* hwm_energy - Obtain energy value
*
* The underlying energy hardware register is 32-bits and is subject to
* overflow. How long before overflow? For example, with an example
* scaling bit shift of 14 bits (see register *PACKAGE_POWER_SKU_UNIT) and
* a power draw of 1000 watts, the 32-bit counter will overflow in
* approximately 4.36 minutes.
*
* Examples:
* 1 watt: (2^32 >> 14) / 1 W / (60 * 60 * 24) secs/day -> 3 days
* 1000 watts: (2^32 >> 14) / 1000 W / 60 secs/min -> 4.36 minutes
*
* The function significantly increases overflow duration (from 4.36
* minutes) by accumulating the energy register into a 'long' as allowed by
* the hwmon API. Using x86_64 128 bit arithmetic (see mul_u64_u32_shr()),
* a 'long' of 63 bits, SF_ENERGY of 1e6 (~20 bits) and
* hwmon->scl_shift_energy of 14 bits we have 57 (63 - 20 + 14) bits before
* energy1_input overflows. This at 1000 W is an overflow duration of 278 years.
*/
static void
hwm_energy(struct hwm_drvdata *ddat, long *energy)
{
struct intel_uncore *uncore = ddat->uncore;
struct i915_hwmon *hwmon = ddat->hwmon;
struct hwm_energy_info *ei = &ddat->ei;
intel_wakeref_t wakeref;
i915_reg_t rgaddr;
u32 reg_val;
if (ddat->gt_n >= 0)
rgaddr = hwmon->rg.energy_status_tile;
else
rgaddr = hwmon->rg.energy_status_all;
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
with_intel_runtime_pm(uncore->rpm, wakeref) {
mutex_lock(&hwmon->hwmon_lock);
reg_val = intel_uncore_read(uncore, rgaddr);
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
if (reg_val >= ei->reg_val_prev)
ei->accum_energy += reg_val - ei->reg_val_prev;
else
ei->accum_energy += UINT_MAX - ei->reg_val_prev + reg_val;
ei->reg_val_prev = reg_val;
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
*energy = mul_u64_u32_shr(ei->accum_energy, SF_ENERGY,
hwmon->scl_shift_energy);
mutex_unlock(&hwmon->hwmon_lock);
}
}
static ssize_t
hwm_power1_max_interval_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
struct i915_hwmon *hwmon = ddat->hwmon;
intel_wakeref_t wakeref;
u32 r, x, y, x_w = 2; /* 2 bits */
u64 tau4, out;
with_intel_runtime_pm(ddat->uncore->rpm, wakeref)
r = intel_uncore_read(ddat->uncore, hwmon->rg.pkg_rapl_limit);
x = REG_FIELD_GET(PKG_PWR_LIM_1_TIME_X, r);
y = REG_FIELD_GET(PKG_PWR_LIM_1_TIME_Y, r);
/*
* tau = 1.x * power(2,y), x = bits(23:22), y = bits(21:17)
* = (4 | x) << (y - 2)
* where (y - 2) ensures a 1.x fixed point representation of 1.x
* However because y can be < 2, we compute
* tau4 = (4 | x) << y
* but add 2 when doing the final right shift to account for units
*/
tau4 = (u64)((1 << x_w) | x) << y;
/* val in hwmon interface units (millisec) */
out = mul_u64_u32_shr(tau4, SF_TIME, hwmon->scl_shift_time + x_w);
return sysfs_emit(buf, "%llu\n", out);
}
static ssize_t
hwm_power1_max_interval_store(struct device *dev,
struct device_attribute *attr,
const char *buf, size_t count)
{
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
struct i915_hwmon *hwmon = ddat->hwmon;
u32 x, y, rxy, x_w = 2; /* 2 bits */
u64 tau4, r, max_win;
unsigned long val;
int ret;
ret = kstrtoul(buf, 0, &val);
if (ret)
return ret;
/*
* Max HW supported tau in '1.x * power(2,y)' format, x = 0, y = 0x12
* The hwmon->scl_shift_time default of 0xa results in a max tau of 256 seconds
*/
#define PKG_MAX_WIN_DEFAULT 0x12ull
/*
* val must be < max in hwmon interface units. The steps below are
* explained in i915_power1_max_interval_show()
*/
r = FIELD_PREP(PKG_MAX_WIN, PKG_MAX_WIN_DEFAULT);
x = REG_FIELD_GET(PKG_MAX_WIN_X, r);
y = REG_FIELD_GET(PKG_MAX_WIN_Y, r);
tau4 = (u64)((1 << x_w) | x) << y;
max_win = mul_u64_u32_shr(tau4, SF_TIME, hwmon->scl_shift_time + x_w);
if (val > max_win)
return -EINVAL;
/* val in hw units */
val = DIV_ROUND_CLOSEST_ULL((u64)val << hwmon->scl_shift_time, SF_TIME);
/* Convert to 1.x * power(2,y) */
if (!val) {
/* Avoid ilog2(0) */
y = 0;
x = 0;
} else {
y = ilog2(val);
/* x = (val - (1 << y)) >> (y - 2); */
x = (val - (1ul << y)) << x_w >> y;
}
rxy = REG_FIELD_PREP(PKG_PWR_LIM_1_TIME_X, x) | REG_FIELD_PREP(PKG_PWR_LIM_1_TIME_Y, y);
hwm_locked_with_pm_intel_uncore_rmw(ddat, hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1_TIME, rxy);
return count;
}
static SENSOR_DEVICE_ATTR(power1_max_interval, 0664,
hwm_power1_max_interval_show,
hwm_power1_max_interval_store, 0);
static struct attribute *hwm_attributes[] = {
&sensor_dev_attr_power1_max_interval.dev_attr.attr,
NULL
};
static umode_t hwm_attributes_visible(struct kobject *kobj,
struct attribute *attr, int index)
{
struct device *dev = kobj_to_dev(kobj);
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
struct i915_hwmon *hwmon = ddat->hwmon;
if (attr == &sensor_dev_attr_power1_max_interval.dev_attr.attr)
return i915_mmio_reg_valid(hwmon->rg.pkg_rapl_limit) ? attr->mode : 0;
return 0;
}
static const struct attribute_group hwm_attrgroup = {
.attrs = hwm_attributes,
.is_visible = hwm_attributes_visible,
};
static const struct attribute_group *hwm_groups[] = {
&hwm_attrgroup,
NULL
};
static const struct hwmon_channel_info * const hwm_info[] = {
HWMON_CHANNEL_INFO(temp, HWMON_T_INPUT),
HWMON_CHANNEL_INFO(in, HWMON_I_INPUT),
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
};
static const struct hwmon_channel_info * const hwm_gt_info[] = {
HWMON_CHANNEL_INFO(energy, HWMON_E_INPUT),
NULL
};
/* I1 is exposed as power_crit or as curr_crit depending on bit 31 */
static int hwm_pcode_read_i1(struct drm_i915_private *i915, u32 *uval)
{
/* Avoid ILLEGAL_SUBCOMMAND "mailbox access failed" warning in snb_pcode_read */
if (IS_DG1(i915) || IS_DG2(i915))
return -ENXIO;
return snb_pcode_read_p(&i915->uncore, PCODE_POWER_SETUP,
POWER_SETUP_SUBCOMMAND_READ_I1, 0, uval);
}
static int hwm_pcode_write_i1(struct drm_i915_private *i915, u32 uval)
{
return snb_pcode_write_p(&i915->uncore, PCODE_POWER_SETUP,
POWER_SETUP_SUBCOMMAND_WRITE_I1, 0, uval);
}
static umode_t
hwm_temp_is_visible(const struct hwm_drvdata *ddat, u32 attr)
{
struct i915_hwmon *hwmon = ddat->hwmon;
if (attr == hwmon_temp_input && i915_mmio_reg_valid(hwmon->rg.pkg_temp))
return 0444;
return 0;
}
static int
hwm_temp_read(struct hwm_drvdata *ddat, u32 attr, long *val)
{
struct i915_hwmon *hwmon = ddat->hwmon;
intel_wakeref_t wakeref;
u32 reg_val;
switch (attr) {
case hwmon_temp_input:
with_intel_runtime_pm(ddat->uncore->rpm, wakeref)
reg_val = intel_uncore_read(ddat->uncore, hwmon->rg.pkg_temp);
/* HW register value is in degrees Celsius, convert to millidegrees. */
*val = REG_FIELD_GET(TEMP_MASK, reg_val) * MILLIDEGREE_PER_DEGREE;
return 0;
default:
return -EOPNOTSUPP;
}
}
static umode_t
hwm_in_is_visible(const struct hwm_drvdata *ddat, u32 attr)
{
struct drm_i915_private *i915 = ddat->uncore->i915;
switch (attr) {
case hwmon_in_input:
return IS_DG1(i915) || IS_DG2(i915) ? 0444 : 0;
default:
return 0;
}
}
static int
hwm_in_read(struct hwm_drvdata *ddat, u32 attr, long *val)
{
struct i915_hwmon *hwmon = ddat->hwmon;
intel_wakeref_t wakeref;
u32 reg_value;
switch (attr) {
case hwmon_in_input:
with_intel_runtime_pm(ddat->uncore->rpm, wakeref)
reg_value = intel_uncore_read(ddat->uncore, hwmon->rg.gt_perf_status);
/* HW register value in units of 2.5 millivolt */
*val = DIV_ROUND_CLOSEST(REG_FIELD_GET(GEN12_VOLTAGE_MASK, reg_value) * 25, 10);
return 0;
default:
return -EOPNOTSUPP;
}
}
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
static umode_t
hwm_power_is_visible(const struct hwm_drvdata *ddat, u32 attr, int chan)
{
struct drm_i915_private *i915 = ddat->uncore->i915;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
struct i915_hwmon *hwmon = ddat->hwmon;
u32 uval;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
switch (attr) {
case hwmon_power_max:
return i915_mmio_reg_valid(hwmon->rg.pkg_rapl_limit) ? 0664 : 0;
case hwmon_power_rated_max:
return i915_mmio_reg_valid(hwmon->rg.pkg_power_sku) ? 0444 : 0;
case hwmon_power_crit:
return (hwm_pcode_read_i1(i915, &uval) ||
!(uval & POWER_SETUP_I1_WATTS)) ? 0 : 0644;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
default:
return 0;
}
}
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
#define PL1_DISABLE 0
/*
* HW allows arbitrary PL1 limits to be set but silently clamps these values to
* "typical but not guaranteed" min/max values in rg.pkg_power_sku. Follow the
* same pattern for sysfs, allow arbitrary PL1 limits to be set but display
* clamped values when read. Write/read I1 also follows the same pattern.
*/
static int
hwm_power_max_read(struct hwm_drvdata *ddat, long *val)
{
struct i915_hwmon *hwmon = ddat->hwmon;
intel_wakeref_t wakeref;
u64 r, min, max;
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
/* Check if PL1 limit is disabled */
with_intel_runtime_pm(ddat->uncore->rpm, wakeref)
r = intel_uncore_read(ddat->uncore, hwmon->rg.pkg_rapl_limit);
if (!(r & PKG_PWR_LIM_1_EN)) {
*val = PL1_DISABLE;
return 0;
}
*val = hwm_field_read_and_scale(ddat,
hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1,
hwmon->scl_shift_power,
SF_POWER);
with_intel_runtime_pm(ddat->uncore->rpm, wakeref)
r = intel_uncore_read64(ddat->uncore, hwmon->rg.pkg_power_sku);
min = REG_FIELD_GET(PKG_MIN_PWR, r);
min = mul_u64_u32_shr(min, SF_POWER, hwmon->scl_shift_power);
max = REG_FIELD_GET(PKG_MAX_PWR, r);
max = mul_u64_u32_shr(max, SF_POWER, hwmon->scl_shift_power);
if (min && max)
*val = clamp_t(u64, *val, min, max);
return 0;
}
static int
hwm_power_max_write(struct hwm_drvdata *ddat, long val)
{
struct i915_hwmon *hwmon = ddat->hwmon;
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
intel_wakeref_t wakeref;
DEFINE_WAIT(wait);
int ret = 0;
u32 nval;
/* Block waiting for GuC reset to complete when needed */
for (;;) {
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
wakeref = intel_runtime_pm_get(ddat->uncore->rpm);
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
mutex_lock(&hwmon->hwmon_lock);
prepare_to_wait(&ddat->waitq, &wait, TASK_INTERRUPTIBLE);
if (!hwmon->ddat.reset_in_progress)
break;
if (signal_pending(current)) {
ret = -EINTR;
break;
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
}
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
mutex_unlock(&hwmon->hwmon_lock);
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
schedule();
}
finish_wait(&ddat->waitq, &wait);
if (ret)
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
goto exit;
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
/* Disable PL1 limit and verify, because the limit cannot be disabled on all platforms */
if (val == PL1_DISABLE) {
intel_uncore_rmw(ddat->uncore, hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1_EN, 0);
nval = intel_uncore_read(ddat->uncore, hwmon->rg.pkg_rapl_limit);
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
if (nval & PKG_PWR_LIM_1_EN)
ret = -ENODEV;
goto exit;
drm/i915/hwmon: Use 0 to designate disabled PL1 power limit On ATSM the PL1 limit is disabled at power up. The previous uapi assumed that the PL1 limit is always enabled and therefore did not have a notion of a disabled PL1 limit. This results in erroneous PL1 limit values when the PL1 limit is disabled. For example at power up, the disabled ATSM PL1 limit was previously shown as 0 which means a low PL1 limit whereas the limit being disabled actually implies a high effective PL1 limit value. To get round this problem, the PL1 limit uapi is expanded to include a special value 0 to designate a disabled PL1 limit. A read value of 0 means that the PL1 power limit is disabled, writing 0 disables the limit. The link between this patch and the bugs mentioned below is as follows: * Because on ATSM the PL1 power limit is disabled on power up and there were no means to enable it, we previously implemented the means to enable the limit when the PL1 hwmon entry (power1_max) was written to. * Now there is a IGT igt@i915_hwmon@hwmon_write which (a) reads orig value from all hwmon sysfs (b) does a bunch of random writes and finally (c) restores the orig value read. On ATSM since the orig value is 0, when the IGT restores the 0 value, the PL1 limit is now enabled with a value of 0. * PL1 limit of 0 implies a low PL1 limit which causes GPU freq to fall to 100 MHz. This causes GuC FW load and several IGT's to start timing out and gives rise to these Intel CI bugs. After this patch, writing 0 would disable the PL1 limit instead of enabling it, avoiding the freq drop issue. v2: Add explanation for bugs mentioned below (Rodrigo) v3: Eliminate race during PL1 disable and verify (Tvrtko) Change return to -ENODEV if verify fails (Tvrtko) Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8062 Link: https://gitlab.freedesktop.org/drm/intel/-/issues/8060 Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230401024146.1826092-1-ashutosh.dixit@intel.com
2023-03-31 19:41:46 -07:00
}
/* Computation in 64-bits to avoid overflow. Round to nearest. */
nval = DIV_ROUND_CLOSEST_ULL((u64)val << hwmon->scl_shift_power, SF_POWER);
nval = PKG_PWR_LIM_1_EN | REG_FIELD_PREP(PKG_PWR_LIM_1, nval);
intel_uncore_rmw(ddat->uncore, hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1_EN | PKG_PWR_LIM_1, nval);
exit:
mutex_unlock(&hwmon->hwmon_lock);
drm/i915/hwmon: Fix locking inversion in sysfs getter In i915 hwmon sysfs getter path we now take a hwmon_lock, then acquire an rpm wakeref. That results in lock inversion: <4> [197.079335] ====================================================== <4> [197.085473] WARNING: possible circular locking dependency detected <4> [197.091611] 6.8.0-rc7-Patchwork_129026v7-gc4dc92fb1152+ #1 Not tainted <4> [197.098096] ------------------------------------------------------ <4> [197.104231] prometheus-node/839 is trying to acquire lock: <4> [197.109680] ffffffff82764d80 (fs_reclaim){+.+.}-{0:0}, at: __kmalloc+0x9a/0x350 <4> [197.116939] but task is already holding lock: <4> [197.122730] ffff88811b772a40 (&hwmon->hwmon_lock){+.+.}-{3:3}, at: hwm_energy+0x4b/0x100 [i915] <4> [197.131543] which lock already depends on the new lock. ... <4> [197.507922] Chain exists of: fs_reclaim --> &gt->reset.mutex --> &hwmon->hwmon_lock <4> [197.518528] Possible unsafe locking scenario: <4> [197.524411] CPU0 CPU1 <4> [197.528916] ---- ---- <4> [197.533418] lock(&hwmon->hwmon_lock); <4> [197.537237] lock(&gt->reset.mutex); <4> [197.543376] lock(&hwmon->hwmon_lock); <4> [197.549682] lock(fs_reclaim); ... <4> [197.632548] Call Trace: <4> [197.634990] <TASK> <4> [197.637088] dump_stack_lvl+0x64/0xb0 <4> [197.640738] check_noncircular+0x15e/0x180 <4> [197.652968] check_prev_add+0xe9/0xce0 <4> [197.656705] __lock_acquire+0x179f/0x2300 <4> [197.660694] lock_acquire+0xd8/0x2d0 <4> [197.673009] fs_reclaim_acquire+0xa1/0xd0 <4> [197.680478] __kmalloc+0x9a/0x350 <4> [197.689063] acpi_ns_internalize_name.part.0+0x4a/0xb0 <4> [197.694170] acpi_ns_get_node_unlocked+0x60/0xf0 <4> [197.720608] acpi_ns_get_node+0x3b/0x60 <4> [197.724428] acpi_get_handle+0x57/0xb0 <4> [197.728164] acpi_has_method+0x20/0x50 <4> [197.731896] acpi_pci_set_power_state+0x43/0x120 <4> [197.736485] pci_power_up+0x24/0x1c0 <4> [197.740047] pci_pm_default_resume_early+0x9/0x30 <4> [197.744725] pci_pm_runtime_resume+0x2d/0x90 <4> [197.753911] __rpm_callback+0x3c/0x110 <4> [197.762586] rpm_callback+0x58/0x70 <4> [197.766064] rpm_resume+0x51e/0x730 <4> [197.769542] rpm_resume+0x267/0x730 <4> [197.773020] rpm_resume+0x267/0x730 <4> [197.776498] rpm_resume+0x267/0x730 <4> [197.779974] __pm_runtime_resume+0x49/0x90 <4> [197.784055] __intel_runtime_pm_get+0x19/0xa0 [i915] <4> [197.789070] hwm_energy+0x55/0x100 [i915] <4> [197.793183] hwm_read+0x9a/0x310 [i915] <4> [197.797124] hwmon_attr_show+0x36/0x120 <4> [197.800946] dev_attr_show+0x15/0x60 <4> [197.804509] sysfs_kf_seq_show+0xb5/0x100 Acquire the wakeref before the lock and hold it as long as the lock is also held. Follow that pattern across the whole source file where similar lock inversion can happen. v2: Keep hardware read under the lock so the whole operation of updating energy from hardware is still atomic (Guenter), - instead, acquire the rpm wakeref before the lock and hold it as long as the lock is held, - use the same aproach for other similar places across the i915_hwmon.c source file (Rodrigo). Fixes: 1b44019a93e2 ("drm/i915/guc: Disable PL1 power limit when loading GuC firmware") Signed-off-by: Janusz Krzysztofik <janusz.krzysztofik@linux.intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Guenter Roeck <linux@roeck-us.net> Cc: <stable@vger.kernel.org> # v6.5+ Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240311203500.518675-2-janusz.krzysztofik@linux.intel.com
2024-03-11 21:34:58 +01:00
intel_runtime_pm_put(ddat->uncore->rpm, wakeref);
return ret;
}
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
static int
hwm_power_read(struct hwm_drvdata *ddat, u32 attr, int chan, long *val)
{
struct i915_hwmon *hwmon = ddat->hwmon;
int ret;
u32 uval;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
switch (attr) {
case hwmon_power_max:
return hwm_power_max_read(ddat, val);
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
case hwmon_power_rated_max:
*val = hwm_field_read_and_scale(ddat,
hwmon->rg.pkg_power_sku,
PKG_PKG_TDP,
hwmon->scl_shift_power,
SF_POWER);
return 0;
case hwmon_power_crit:
ret = hwm_pcode_read_i1(ddat->uncore->i915, &uval);
if (ret)
return ret;
if (!(uval & POWER_SETUP_I1_WATTS))
return -ENODEV;
*val = mul_u64_u32_shr(REG_FIELD_GET(POWER_SETUP_I1_DATA_MASK, uval),
SF_POWER, POWER_SETUP_I1_SHIFT);
return 0;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
default:
return -EOPNOTSUPP;
}
}
static int
hwm_power_write(struct hwm_drvdata *ddat, u32 attr, int chan, long val)
{
u32 uval;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
switch (attr) {
case hwmon_power_max:
return hwm_power_max_write(ddat, val);
case hwmon_power_crit:
uval = DIV_ROUND_CLOSEST_ULL(val << POWER_SETUP_I1_SHIFT, SF_POWER);
return hwm_pcode_write_i1(ddat->uncore->i915, uval);
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
default:
return -EOPNOTSUPP;
}
}
void i915_hwmon_power_max_disable(struct drm_i915_private *i915, bool *old)
{
struct i915_hwmon *hwmon = i915->hwmon;
u32 r;
if (!hwmon || !i915_mmio_reg_valid(hwmon->rg.pkg_rapl_limit))
return;
mutex_lock(&hwmon->hwmon_lock);
hwmon->ddat.reset_in_progress = true;
r = intel_uncore_rmw(hwmon->ddat.uncore, hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1_EN, 0);
*old = !!(r & PKG_PWR_LIM_1_EN);
mutex_unlock(&hwmon->hwmon_lock);
}
void i915_hwmon_power_max_restore(struct drm_i915_private *i915, bool old)
{
struct i915_hwmon *hwmon = i915->hwmon;
if (!hwmon || !i915_mmio_reg_valid(hwmon->rg.pkg_rapl_limit))
return;
mutex_lock(&hwmon->hwmon_lock);
intel_uncore_rmw(hwmon->ddat.uncore, hwmon->rg.pkg_rapl_limit,
PKG_PWR_LIM_1_EN, old ? PKG_PWR_LIM_1_EN : 0);
hwmon->ddat.reset_in_progress = false;
wake_up_all(&hwmon->ddat.waitq);
mutex_unlock(&hwmon->hwmon_lock);
}
static umode_t
hwm_energy_is_visible(const struct hwm_drvdata *ddat, u32 attr)
{
struct i915_hwmon *hwmon = ddat->hwmon;
i915_reg_t rgaddr;
switch (attr) {
case hwmon_energy_input:
if (ddat->gt_n >= 0)
rgaddr = hwmon->rg.energy_status_tile;
else
rgaddr = hwmon->rg.energy_status_all;
return i915_mmio_reg_valid(rgaddr) ? 0444 : 0;
default:
return 0;
}
}
static int
hwm_energy_read(struct hwm_drvdata *ddat, u32 attr, long *val)
{
switch (attr) {
case hwmon_energy_input:
hwm_energy(ddat, val);
return 0;
default:
return -EOPNOTSUPP;
}
}
static umode_t
hwm_curr_is_visible(const struct hwm_drvdata *ddat, u32 attr)
{
struct drm_i915_private *i915 = ddat->uncore->i915;
u32 uval;
switch (attr) {
case hwmon_curr_crit:
return (hwm_pcode_read_i1(i915, &uval) ||
(uval & POWER_SETUP_I1_WATTS)) ? 0 : 0644;
default:
return 0;
}
}
static int
hwm_curr_read(struct hwm_drvdata *ddat, u32 attr, long *val)
{
int ret;
u32 uval;
switch (attr) {
case hwmon_curr_crit:
ret = hwm_pcode_read_i1(ddat->uncore->i915, &uval);
if (ret)
return ret;
if (uval & POWER_SETUP_I1_WATTS)
return -ENODEV;
*val = mul_u64_u32_shr(REG_FIELD_GET(POWER_SETUP_I1_DATA_MASK, uval),
SF_CURR, POWER_SETUP_I1_SHIFT);
return 0;
default:
return -EOPNOTSUPP;
}
}
static int
hwm_curr_write(struct hwm_drvdata *ddat, u32 attr, long val)
{
u32 uval;
switch (attr) {
case hwmon_curr_crit:
uval = DIV_ROUND_CLOSEST_ULL(val << POWER_SETUP_I1_SHIFT, SF_CURR);
return hwm_pcode_write_i1(ddat->uncore->i915, uval);
default:
return -EOPNOTSUPP;
}
}
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_ULL(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)
{
struct hwm_drvdata *ddat = (struct hwm_drvdata *)drvdata;
switch (type) {
case hwmon_temp:
return hwm_temp_is_visible(ddat, attr);
case hwmon_in:
return hwm_in_is_visible(ddat, attr);
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
case hwmon_power:
return hwm_power_is_visible(ddat, attr, channel);
case hwmon_energy:
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;
}
}
static int
hwm_read(struct device *dev, enum hwmon_sensor_types type, u32 attr,
int channel, long *val)
{
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
switch (type) {
case hwmon_temp:
return hwm_temp_read(ddat, attr, val);
case hwmon_in:
return hwm_in_read(ddat, attr, val);
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
case hwmon_power:
return hwm_power_read(ddat, attr, channel, val);
case hwmon_energy:
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;
}
}
static int
hwm_write(struct device *dev, enum hwmon_sensor_types type, u32 attr,
int channel, long val)
{
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
switch (type) {
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
case hwmon_power:
return hwm_power_write(ddat, attr, channel, val);
case hwmon_curr:
return hwm_curr_write(ddat, attr, val);
default:
return -EOPNOTSUPP;
}
}
static const struct hwmon_ops hwm_ops = {
.is_visible = hwm_is_visible,
.read = hwm_read,
.write = hwm_write,
};
static const struct hwmon_chip_info hwm_chip_info = {
.ops = &hwm_ops,
.info = hwm_info,
};
static umode_t
hwm_gt_is_visible(const void *drvdata, enum hwmon_sensor_types type,
u32 attr, int channel)
{
struct hwm_drvdata *ddat = (struct hwm_drvdata *)drvdata;
switch (type) {
case hwmon_energy:
return hwm_energy_is_visible(ddat, attr);
default:
return 0;
}
}
static int
hwm_gt_read(struct device *dev, enum hwmon_sensor_types type, u32 attr,
int channel, long *val)
{
struct hwm_drvdata *ddat = dev_get_drvdata(dev);
switch (type) {
case hwmon_energy:
return hwm_energy_read(ddat, attr, val);
default:
return -EOPNOTSUPP;
}
}
static const struct hwmon_ops hwm_gt_ops = {
.is_visible = hwm_gt_is_visible,
.read = hwm_gt_read,
};
static const struct hwmon_chip_info hwm_gt_chip_info = {
.ops = &hwm_gt_ops,
.info = hwm_gt_info,
};
static void
hwm_get_preregistration_info(struct drm_i915_private *i915)
{
struct i915_hwmon *hwmon = i915->hwmon;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
struct intel_uncore *uncore = &i915->uncore;
struct hwm_drvdata *ddat = &hwmon->ddat;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
intel_wakeref_t wakeref;
u32 val_sku_unit = 0;
struct intel_gt *gt;
long energy;
int i;
/* Available for all Gen12+/dGfx */
hwmon->rg.gt_perf_status = GEN12_RPSTAT1;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
if (IS_DG1(i915) || IS_DG2(i915)) {
hwmon->rg.pkg_temp = PCU_PACKAGE_TEMPERATURE;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
hwmon->rg.pkg_power_sku_unit = PCU_PACKAGE_POWER_SKU_UNIT;
hwmon->rg.pkg_power_sku = PCU_PACKAGE_POWER_SKU;
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;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
} else {
hwmon->rg.pkg_temp = INVALID_MMIO_REG;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
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;
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
}
with_intel_runtime_pm(uncore->rpm, wakeref) {
/*
* The contents of register hwmon->rg.pkg_power_sku_unit do not change,
* so read it once and store the shift values.
*/
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();
}
drm/i915/hwmon: Power PL1 limit and TDP setting Use i915 HWMON to display/modify dGfx power PL1 limit and TDP setting. v2: - Fix review comments (Ashutosh) - Do not restore power1_max upon module unload/load sequence because on production systems modules are always loaded and not unloaded/reloaded (Ashutosh) - Fix review comments (Jani) - Remove endianness conversion (Ashutosh) v3: Add power1_rated_max (Ashutosh) v4: - Use macro HWMON_CHANNEL_INFO to define power channel (Guenter) - Update the date and kernel version in Documentation (Badal) v5: Use hwm_ prefix for static functions (Ashutosh) v6: Fix review comments (Ashutosh) v7: - Define PCU_PACKAGE_POWER_SKU for DG1,DG2 and move PKG_PKG_TDP to intel_mchbar_regs.h (Anshuman) - KernelVersion: 6.2, Date: February 2023 in doc (Tvrtko) v8: Change contact to intel-gfx (Rodrigo) Minor change to val_sku_unit init (Andi) Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Dale B Stimson <dale.b.stimson@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Signed-off-by: Riana Tauro <riana.tauro@intel.com> Signed-off-by: Badal Nilawar <badal.nilawar@intel.com> Acked-by: Guenter Roeck <linux@roeck-us.net> Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Anshuman Gupta <anshuman.gupta@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20221013154526.2105579-4-ashutosh.dixit@intel.com
2022-10-13 08:45:22 -07:00
}
hwmon->scl_shift_power = REG_FIELD_GET(PKG_PWR_UNIT, val_sku_unit);
hwmon->scl_shift_energy = REG_FIELD_GET(PKG_ENERGY_UNIT, val_sku_unit);
hwmon->scl_shift_time = REG_FIELD_GET(PKG_TIME_UNIT, val_sku_unit);
/*
* Initialize 'struct hwm_energy_info', i.e. set fields to the
* first value of the energy register read
*/
if (i915_mmio_reg_valid(hwmon->rg.energy_status_all))
hwm_energy(ddat, &energy);
if (i915_mmio_reg_valid(hwmon->rg.energy_status_tile)) {
for_each_gt(gt, i915, i)
hwm_energy(&hwmon->ddat_gt[i], &energy);
}
}
void i915_hwmon_register(struct drm_i915_private *i915)
{
struct device *dev = i915->drm.dev;
struct i915_hwmon *hwmon;
struct device *hwmon_dev;
struct hwm_drvdata *ddat;
struct hwm_drvdata *ddat_gt;
struct intel_gt *gt;
int i;
/* hwmon is available only for dGfx */
if (!IS_DGFX(i915))
return;
drm/i915/hwmon: Get rid of devm When both hwmon and hwmon drvdata (on which hwmon depends) are device managed resources, the expectation, on device unbind, is that hwmon will be released before drvdata. However, in i915 there are two separate code paths, which both release either drvdata or hwmon and either can be released before the other. These code paths (for device unbind) are as follows (see also the bug referenced below): Call Trace: release_nodes+0x11/0x70 devres_release_group+0xb2/0x110 component_unbind_all+0x8d/0xa0 component_del+0xa5/0x140 intel_pxp_tee_component_fini+0x29/0x40 [i915] intel_pxp_fini+0x33/0x80 [i915] i915_driver_remove+0x4c/0x120 [i915] i915_pci_remove+0x19/0x30 [i915] pci_device_remove+0x32/0xa0 device_release_driver_internal+0x19c/0x200 unbind_store+0x9c/0xb0 and Call Trace: release_nodes+0x11/0x70 devres_release_all+0x8a/0xc0 device_unbind_cleanup+0x9/0x70 device_release_driver_internal+0x1c1/0x200 unbind_store+0x9c/0xb0 This means that in i915, if use devm, we cannot gurantee that hwmon will always be released before drvdata. Which means that we have a uaf if hwmon sysfs is accessed when drvdata has been released but hwmon hasn't. The only way out of this seems to be do get rid of devm_ and release/free everything explicitly during device unbind. v2: Change commit message and other minor code changes v3: Cleanup from i915_hwmon_register on error (Armin Wolf) v4: Eliminate potential static analyzer warning (Rodrigo) Eliminate fetch_and_zero (Jani) v5: Restore previous logic for ddat_gt->hwmon_dev error return (Andi) Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/10366 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240417145646.793223-1-ashutosh.dixit@intel.com
2024-04-17 07:56:46 -07:00
hwmon = kzalloc(sizeof(*hwmon), GFP_KERNEL);
if (!hwmon)
return;
i915->hwmon = hwmon;
mutex_init(&hwmon->hwmon_lock);
ddat = &hwmon->ddat;
ddat->hwmon = hwmon;
ddat->uncore = &i915->uncore;
snprintf(ddat->name, sizeof(ddat->name), "i915");
ddat->gt_n = -1;
init_waitqueue_head(&ddat->waitq);
for_each_gt(gt, i915, i) {
ddat_gt = hwmon->ddat_gt + i;
ddat_gt->hwmon = hwmon;
ddat_gt->uncore = gt->uncore;
snprintf(ddat_gt->name, sizeof(ddat_gt->name), "i915_gt%u", i);
ddat_gt->gt_n = i;
}
hwm_get_preregistration_info(i915);
/* hwmon_dev points to device hwmon<i> */
drm/i915/hwmon: Get rid of devm When both hwmon and hwmon drvdata (on which hwmon depends) are device managed resources, the expectation, on device unbind, is that hwmon will be released before drvdata. However, in i915 there are two separate code paths, which both release either drvdata or hwmon and either can be released before the other. These code paths (for device unbind) are as follows (see also the bug referenced below): Call Trace: release_nodes+0x11/0x70 devres_release_group+0xb2/0x110 component_unbind_all+0x8d/0xa0 component_del+0xa5/0x140 intel_pxp_tee_component_fini+0x29/0x40 [i915] intel_pxp_fini+0x33/0x80 [i915] i915_driver_remove+0x4c/0x120 [i915] i915_pci_remove+0x19/0x30 [i915] pci_device_remove+0x32/0xa0 device_release_driver_internal+0x19c/0x200 unbind_store+0x9c/0xb0 and Call Trace: release_nodes+0x11/0x70 devres_release_all+0x8a/0xc0 device_unbind_cleanup+0x9/0x70 device_release_driver_internal+0x1c1/0x200 unbind_store+0x9c/0xb0 This means that in i915, if use devm, we cannot gurantee that hwmon will always be released before drvdata. Which means that we have a uaf if hwmon sysfs is accessed when drvdata has been released but hwmon hasn't. The only way out of this seems to be do get rid of devm_ and release/free everything explicitly during device unbind. v2: Change commit message and other minor code changes v3: Cleanup from i915_hwmon_register on error (Armin Wolf) v4: Eliminate potential static analyzer warning (Rodrigo) Eliminate fetch_and_zero (Jani) v5: Restore previous logic for ddat_gt->hwmon_dev error return (Andi) Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/10366 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240417145646.793223-1-ashutosh.dixit@intel.com
2024-04-17 07:56:46 -07:00
hwmon_dev = hwmon_device_register_with_info(dev, ddat->name,
ddat,
&hwm_chip_info,
hwm_groups);
if (IS_ERR(hwmon_dev))
goto err;
ddat->hwmon_dev = hwmon_dev;
for_each_gt(gt, i915, i) {
ddat_gt = hwmon->ddat_gt + i;
/*
* Create per-gt directories only if a per-gt attribute is
* visible. Currently this is only energy
*/
if (!hwm_gt_is_visible(ddat_gt, hwmon_energy, hwmon_energy_input, 0))
continue;
drm/i915/hwmon: Get rid of devm When both hwmon and hwmon drvdata (on which hwmon depends) are device managed resources, the expectation, on device unbind, is that hwmon will be released before drvdata. However, in i915 there are two separate code paths, which both release either drvdata or hwmon and either can be released before the other. These code paths (for device unbind) are as follows (see also the bug referenced below): Call Trace: release_nodes+0x11/0x70 devres_release_group+0xb2/0x110 component_unbind_all+0x8d/0xa0 component_del+0xa5/0x140 intel_pxp_tee_component_fini+0x29/0x40 [i915] intel_pxp_fini+0x33/0x80 [i915] i915_driver_remove+0x4c/0x120 [i915] i915_pci_remove+0x19/0x30 [i915] pci_device_remove+0x32/0xa0 device_release_driver_internal+0x19c/0x200 unbind_store+0x9c/0xb0 and Call Trace: release_nodes+0x11/0x70 devres_release_all+0x8a/0xc0 device_unbind_cleanup+0x9/0x70 device_release_driver_internal+0x1c1/0x200 unbind_store+0x9c/0xb0 This means that in i915, if use devm, we cannot gurantee that hwmon will always be released before drvdata. Which means that we have a uaf if hwmon sysfs is accessed when drvdata has been released but hwmon hasn't. The only way out of this seems to be do get rid of devm_ and release/free everything explicitly during device unbind. v2: Change commit message and other minor code changes v3: Cleanup from i915_hwmon_register on error (Armin Wolf) v4: Eliminate potential static analyzer warning (Rodrigo) Eliminate fetch_and_zero (Jani) v5: Restore previous logic for ddat_gt->hwmon_dev error return (Andi) Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/10366 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240417145646.793223-1-ashutosh.dixit@intel.com
2024-04-17 07:56:46 -07:00
hwmon_dev = hwmon_device_register_with_info(dev, ddat_gt->name,
ddat_gt,
&hwm_gt_chip_info,
NULL);
if (!IS_ERR(hwmon_dev))
ddat_gt->hwmon_dev = hwmon_dev;
}
drm/i915/hwmon: Get rid of devm When both hwmon and hwmon drvdata (on which hwmon depends) are device managed resources, the expectation, on device unbind, is that hwmon will be released before drvdata. However, in i915 there are two separate code paths, which both release either drvdata or hwmon and either can be released before the other. These code paths (for device unbind) are as follows (see also the bug referenced below): Call Trace: release_nodes+0x11/0x70 devres_release_group+0xb2/0x110 component_unbind_all+0x8d/0xa0 component_del+0xa5/0x140 intel_pxp_tee_component_fini+0x29/0x40 [i915] intel_pxp_fini+0x33/0x80 [i915] i915_driver_remove+0x4c/0x120 [i915] i915_pci_remove+0x19/0x30 [i915] pci_device_remove+0x32/0xa0 device_release_driver_internal+0x19c/0x200 unbind_store+0x9c/0xb0 and Call Trace: release_nodes+0x11/0x70 devres_release_all+0x8a/0xc0 device_unbind_cleanup+0x9/0x70 device_release_driver_internal+0x1c1/0x200 unbind_store+0x9c/0xb0 This means that in i915, if use devm, we cannot gurantee that hwmon will always be released before drvdata. Which means that we have a uaf if hwmon sysfs is accessed when drvdata has been released but hwmon hasn't. The only way out of this seems to be do get rid of devm_ and release/free everything explicitly during device unbind. v2: Change commit message and other minor code changes v3: Cleanup from i915_hwmon_register on error (Armin Wolf) v4: Eliminate potential static analyzer warning (Rodrigo) Eliminate fetch_and_zero (Jani) v5: Restore previous logic for ddat_gt->hwmon_dev error return (Andi) Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/10366 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240417145646.793223-1-ashutosh.dixit@intel.com
2024-04-17 07:56:46 -07:00
return;
err:
i915_hwmon_unregister(i915);
}
void i915_hwmon_unregister(struct drm_i915_private *i915)
{
drm/i915/hwmon: Get rid of devm When both hwmon and hwmon drvdata (on which hwmon depends) are device managed resources, the expectation, on device unbind, is that hwmon will be released before drvdata. However, in i915 there are two separate code paths, which both release either drvdata or hwmon and either can be released before the other. These code paths (for device unbind) are as follows (see also the bug referenced below): Call Trace: release_nodes+0x11/0x70 devres_release_group+0xb2/0x110 component_unbind_all+0x8d/0xa0 component_del+0xa5/0x140 intel_pxp_tee_component_fini+0x29/0x40 [i915] intel_pxp_fini+0x33/0x80 [i915] i915_driver_remove+0x4c/0x120 [i915] i915_pci_remove+0x19/0x30 [i915] pci_device_remove+0x32/0xa0 device_release_driver_internal+0x19c/0x200 unbind_store+0x9c/0xb0 and Call Trace: release_nodes+0x11/0x70 devres_release_all+0x8a/0xc0 device_unbind_cleanup+0x9/0x70 device_release_driver_internal+0x1c1/0x200 unbind_store+0x9c/0xb0 This means that in i915, if use devm, we cannot gurantee that hwmon will always be released before drvdata. Which means that we have a uaf if hwmon sysfs is accessed when drvdata has been released but hwmon hasn't. The only way out of this seems to be do get rid of devm_ and release/free everything explicitly during device unbind. v2: Change commit message and other minor code changes v3: Cleanup from i915_hwmon_register on error (Armin Wolf) v4: Eliminate potential static analyzer warning (Rodrigo) Eliminate fetch_and_zero (Jani) v5: Restore previous logic for ddat_gt->hwmon_dev error return (Andi) Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/10366 Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20240417145646.793223-1-ashutosh.dixit@intel.com
2024-04-17 07:56:46 -07:00
struct i915_hwmon *hwmon = i915->hwmon;
struct intel_gt *gt;
int i;
if (!hwmon)
return;
for_each_gt(gt, i915, i)
if (hwmon->ddat_gt[i].hwmon_dev)
hwmon_device_unregister(hwmon->ddat_gt[i].hwmon_dev);
if (hwmon->ddat.hwmon_dev)
hwmon_device_unregister(hwmon->ddat.hwmon_dev);
mutex_destroy(&hwmon->hwmon_lock);
kfree(i915->hwmon);
i915->hwmon = NULL;
}