RE: [RFC 2/2] drm/amd/pm: Add support for reacting to platform profile notification

["Limonciello, Mario" <Mario.Limonciello@amd.com>]

[-- Attachment #1: Type: text/plain, Size: 10788 bytes --]

[Public]

> I don't think we want to force the performance level.  This interface
forces various fixed clock configurations for debugging and profiling.
Ah got it.

>I think what we'd want to select here is the power profile (see
amdgpu_set_pp_power_profile_mode()).  For this interface you can
select various profiles (BOOTUP_DEFAULT, 3D_FULL_SCREEN, POWER_SAVING,
VIDEO, VR, COMPUTE, etc.).  These still use dynamic power management,
but they adjust the heuristics used by the GPU to select power states
so the GPU performance ramps up/down more or less aggressively.

Which profile mapping you think make sense?
My guess would be:
"BOOTUP_DEFAULT" for balanced
"POWER_SAVING" for low-power
"3D_FULL_SCREEN" for performance

Since recently we removed that interface for YC, and some earlier APUs don't do as much with it.
So I wonder if this is only really valuable to do this callback for !APU.

> I feel it's better to leave to platform vendors. For ex: for APU cases they may have implementations in which their BIOSes talk to PMFW and this might be driving something else here.

Even if changing the heuristic for workload as Alex suggested?

> Also, not sure how to handle a case like, say a laptop with Intel CPU and AMD dgpu.

I was actually thinking this approach where there are platform profile callbacks is best because of that specifically.
It would allow an Intel CPU system to have a platform profile driver implemented by the OEM, but then still notify amdgpu dGPU to tune for power saving or performance workload.

From: Lazar, Lijo <Lijo.Lazar@amd.com>
Sent: Wednesday, November 10, 2021 10:05
To: Alex Deucher <alexdeucher@gmail.com>; Limonciello, Mario <Mario.Limonciello@amd.com>
Cc: amd-gfx list <amd-gfx@lists.freedesktop.org>
Subject: Re: [RFC 2/2] drm/amd/pm: Add support for reacting to platform profile notification


[Public]

I feel it's better to leave to platform vendors. For ex: for APU cases they may have implementations in which their BIOSes talk to PMFW and this might be driving something else here.

Also, not sure how to handle a case like, say a laptop with Intel CPU and AMD dgpu.

Thanks,
Lijo

________________________________
From: amd-gfx <amd-gfx-bounces@lists.freedesktop.org<mailto:amd-gfx-bounces@lists.freedesktop.org>> on behalf of Alex Deucher <alexdeucher@gmail.com<mailto:alexdeucher@gmail.com>>
Sent: Wednesday, 10 November 2021, 8:44 pm
To: Limonciello, Mario
Cc: amd-gfx list
Subject: Re: [RFC 2/2] drm/amd/pm: Add support for reacting to platform profile notification

On Wed, Nov 10, 2021 at 1:24 AM Mario Limonciello
<mario.limonciello@amd.com<mailto:mario.limonciello@amd.com>> wrote:
>
> Various drivers provide platform profile support to let users set a hint
> in their GUI whether they want to run in a high performance, low battery
> life or balanced configuration.
>
> Drivers that provide this typically work with the firmware on their system
> to configure hardware.  In the case of AMDGPU however, the notification
> path doesn't come through firmware and can instead be provided directly
> to the driver from a notification chain.
>
> Use the information of the newly selected profile to tweak
> `dpm_force_performance_level` to that profile IFF the user hasn't manually
> selected `manual` or any other `profile_*` options.

I don't think we want to force the performance level.  This interface
forces various fixed clock configurations for debugging and profiling.
I think what we'd want to select here is the power profile (see
amdgpu_set_pp_power_profile_mode()).  For this interface you can
select various profiles (BOOTUP_DEFAULT, 3D_FULL_SCREEN, POWER_SAVING,
VIDEO, VR, COMPUTE, etc.).  These still use dynamic power management,
but they adjust the heuristics used by the GPU to select power states
so the GPU performance ramps up/down more or less aggressively.

Alex

>
> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com<mailto:mario.limonciello@amd.com>>
> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu.h |   3 +
>  drivers/gpu/drm/amd/pm/amdgpu_pm.c  | 105 +++++++++++++++++++++++-----
>  2 files changed, 90 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
> index b85b67a88a3d..27b0be23b6ac 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
> @@ -1097,6 +1097,9 @@ struct amdgpu_device {
>
>         struct amdgpu_reset_control     *reset_cntl;
>         uint32_t                        ip_versions[HW_ID_MAX][HWIP_MAX_INSTANCE];
> +
> +       /* platform profile notifications */
> +       struct notifier_block           platform_profile_notifier;
>  };
>
>  static inline struct amdgpu_device *drm_to_adev(struct drm_device *ddev)
> diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
> index 41472ed99253..33fc52b90d4c 100644
> --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c
> +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c
> @@ -32,6 +32,7 @@
>  #include <linux/hwmon.h>
>  #include <linux/hwmon-sysfs.h>
>  #include <linux/nospec.h>
> +#include <linux/platform_profile.h>
>  #include <linux/pm_runtime.h>
>  #include <asm/processor.h>
>  #include "hwmgr.h"
> @@ -200,6 +201,33 @@ static ssize_t amdgpu_set_power_dpm_state(struct device *dev,
>         return count;
>  }
>
> +static int amdgpu_get_forced_level(struct device *dev, enum amd_dpm_forced_level *level)
> +{
> +       struct drm_device *ddev = dev_get_drvdata(dev);
> +       struct amdgpu_device *adev = drm_to_adev(ddev);
> +       int ret;
> +
> +       if (amdgpu_in_reset(adev))
> +               return -EPERM;
> +       if (adev->in_suspend && !adev->in_runpm)
> +               return -EPERM;
> +
> +       ret = pm_runtime_get_sync(ddev->dev);
> +       if (ret < 0) {
> +               pm_runtime_put_autosuspend(ddev->dev);
> +               return ret;
> +       }
> +
> +       if (adev->powerplay.pp_funcs->get_performance_level)
> +               *level = amdgpu_dpm_get_performance_level(adev);
> +       else
> +               *level = adev->pm.dpm.forced_level;
> +
> +       pm_runtime_mark_last_busy(ddev->dev);
> +       pm_runtime_put_autosuspend(ddev->dev);
> +
> +       return 0;
> +}
>
>  /**
>   * DOC: power_dpm_force_performance_level
> @@ -264,29 +292,13 @@ static ssize_t amdgpu_get_power_dpm_force_performance_level(struct device *dev,
>                                                             struct device_attribute *attr,
>                                                             char *buf)
>  {
> -       struct drm_device *ddev = dev_get_drvdata(dev);
> -       struct amdgpu_device *adev = drm_to_adev(ddev);
>         enum amd_dpm_forced_level level = 0xff;
>         int ret;
>
> -       if (amdgpu_in_reset(adev))
> -               return -EPERM;
> -       if (adev->in_suspend && !adev->in_runpm)
> -               return -EPERM;
> +       ret = amdgpu_get_forced_level(dev, &level);
>
> -       ret = pm_runtime_get_sync(ddev->dev);
> -       if (ret < 0) {
> -               pm_runtime_put_autosuspend(ddev->dev);
> +       if (ret < 0)
>                 return ret;
> -       }
> -
> -       if (adev->powerplay.pp_funcs->get_performance_level)
> -               level = amdgpu_dpm_get_performance_level(adev);
> -       else
> -               level = adev->pm.dpm.forced_level;
> -
> -       pm_runtime_mark_last_busy(ddev->dev);
> -       pm_runtime_put_autosuspend(ddev->dev);
>
>         return sysfs_emit(buf, "%s\n",
>                           (level == AMD_DPM_FORCED_LEVEL_AUTO) ? "auto" :
> @@ -405,6 +417,59 @@ static ssize_t amdgpu_set_power_dpm_force_performance_level(struct device *dev,
>         return count;
>  }
>
> +static void amdgpu_update_profile(struct device *dev, enum platform_profile_option *profile)
> +{
> +       enum amd_dpm_forced_level level;
> +       const char *str;
> +       int ret;
> +
> +       ret = amdgpu_get_forced_level(dev, &level);
> +       if (ret < 0)
> +               return;
> +
> +       /* only update profile if we're in fixed modes right now that need updating */
> +       switch (level) {
> +       case AMD_DPM_FORCED_LEVEL_LOW:
> +               if (*profile < PLATFORM_PROFILE_BALANCED)
> +                       return;
> +               break;
> +       case AMD_DPM_FORCED_LEVEL_HIGH:
> +               if (*profile > PLATFORM_PROFILE_BALANCED)
> +                       return;
> +               break;
> +       case AMD_DPM_FORCED_LEVEL_AUTO:
> +               if (*profile == PLATFORM_PROFILE_BALANCED)
> +                       return;
> +               break;
> +       default:
> +               dev_dbg(dev, "refusing to update amdgpu profile from %d\n", level);
> +               return;
> +       }
> +       if (*profile > PLATFORM_PROFILE_BALANCED)
> +               str = "high";
> +       else if (*profile < PLATFORM_PROFILE_BALANCED)
> +               str = "low";
> +       else
> +               str = "auto";
> +
> +       dev_dbg(dev, "updating platform profile to %s\n", str);
> +       amdgpu_set_power_dpm_force_performance_level(dev, NULL, str, 0);
> +}
> +
> +static int amdgpu_platform_profile_notifier_call(struct notifier_block *nb,
> +                                                 unsigned long action, void *data)
> +{
> +       if (action == PLATFORM_PROFILE_CHANGED) {
> +               enum platform_profile_option *profile = data;
> +               struct amdgpu_device *adev;
> +
> +               adev = container_of(nb, struct amdgpu_device, platform_profile_notifier);
> +               amdgpu_update_profile(adev->dev, profile);
> +       }
> +
> +       return NOTIFY_OK;
> +}
> +
>  static ssize_t amdgpu_get_pp_num_states(struct device *dev,
>                 struct device_attribute *attr,
>                 char *buf)
> @@ -3506,6 +3571,9 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
>         if (ret)
>                 return ret;
>
> +       adev->platform_profile_notifier.notifier_call = amdgpu_platform_profile_notifier_call;
> +       platform_profile_register_notifier(&adev->platform_profile_notifier);
> +
>         adev->pm.sysfs_initialized = true;
>
>         return 0;
> @@ -3519,6 +3587,7 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
>         if (adev->pm.int_hwmon_dev)
>                 hwmon_device_unregister(adev->pm.int_hwmon_dev);
>
> +       platform_profile_unregister_notifier(&adev->platform_profile_notifier);
>         amdgpu_device_attr_remove_groups(adev, &adev->pm.pm_attr_list);
>  }
>
> --
> 2.25.1
>


[-- Attachment #2: Type: text/html, Size: 24509 bytes --]