RE: [PATCH v2] drm/amdgpu: Update message IDs to PMFW to correctly gather GFXOFF residency logs

Public

> -----Original Message-----
> From: amd-gfx <[email protected]> On Behalf Of Fares
> Soliman
> Sent: Monday, July 13, 2026 9:26 AM
> To: [email protected]
> Cc: Liang, Richard qi <[email protected]>; Soliman, Fares
> <[email protected]>
> Subject: [PATCH v2] drm/amdgpu: Update message IDs to PMFW to correctly
> gather GFXOFF residency logs
>
> Updates PPSMC_MSGs and set/get functions for gathering GFXOFF logs on
> Van Gogh. Logs are now gathered live rather than starting then stopping
> logging and reading an average value afterwards. This is in accordance to
> changes made in PMFW.
>
> In regards to messageID 0x52, the old interface uses a start/stop parameter,
> and the new one doesn't. The firmware is checked to determine which method
> to use.
>
> v2: added firmware guard to new interface, old interface kept as fallback
>
> Signed-off-by: Fares Soliman <[email protected]>

Acked-by: Alex Deucher <[email protected]>

> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c   |  4 +-
>  drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h |  2 +-
>  .../pm/swsmu/inc/pmfw_if/smu_v11_5_ppsmc.h    |  6 ++-
>  drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h  |  4 +-
> .../gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c  | 50
> +++++++++++++++----
>  5 files changed, 50 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> index 5c4d4ff001ea..46514a1faf1f 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> @@ -1320,8 +1320,8 @@ static ssize_t amdgpu_debugfs_gpr_read(struct file
> *f, char __user *buf,
>   * @size: Number of bytes to read
>   * @pos:  Offset to seek to
>   *
> - * Read the last residency value logged. It doesn't auto update, one needs to
> - * stop logging before getting the current value.
> + * Read a live GFXOFF residency sample from firmware. One needs to
> + start logging
> + * before getting the current value.
>   */
>  static ssize_t amdgpu_debugfs_gfxoff_residency_read(struct file *f, char
> __user *buf,
>                                                   size_t size, loff_t *pos)
> diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
> b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
> index f8fd93999617..d94e3dcf7f9a 100644
> --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
> +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h
> @@ -1333,7 +1333,7 @@ struct pptable_funcs {
>       u32 (*set_gfx_off_residency)(struct smu_context *smu, bool start);
>
>       /**
> -      * @get_gfx_off_residency: Average GFXOFF residency % during the
> logging interval
> +      * @get_gfx_off_residency: Live GFXOFF residency percentage
>        */
>       u32 (*get_gfx_off_residency)(struct smu_context *smu, uint32_t
> *residency);
>
> diff --git
> a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v11_5_ppsmc.h
> b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v11_5_ppsmc.h
> index 7471e2df2828..4206514765cd 100644
> --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v11_5_ppsmc.h
> +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v11_5_ppsmc.h
> @@ -110,8 +110,10 @@
>  #define PPSMC_MSG_GetSlowPPTLimit                      0x4C
>  #define PPSMC_MSG_GetGfxOffStatus                   0x50
>  #define PPSMC_MSG_GetGfxOffEntryCount                       0x51
> -#define PPSMC_MSG_LogGfxOffResidency                0x52
> -#define PPSMC_Message_Count                            0x53
> +#define PPSMC_MSG_GfxOffResidencyLogReadSample              0x52
> +#define PPSMC_MSG_StopGfxOffResidencyLogging          0x53
> +#define PPSMC_MSG_StartGfxOffResidencyLogging                 0x56
> +#define PPSMC_Message_Count                            0x57
>
>  //Argument for PPSMC_MSG_GfxDeviceDriverReset  enum { diff --git
> a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
> b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
> index 636ff90923d9..acf03838d49d 100644
> --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
> +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h
> @@ -252,7 +252,9 @@
>       __SMU_DUMMY_MAP(DriverMode2Reset), \
>       __SMU_DUMMY_MAP(GetGfxOffStatus),                \
>       __SMU_DUMMY_MAP(GetGfxOffEntryCount),            \
> -     __SMU_DUMMY_MAP(LogGfxOffResidency),                    \
> +     __SMU_DUMMY_MAP(StartGfxOffResidencyLogging),           \
> +     __SMU_DUMMY_MAP(GfxOffResidencyLogReadSample),
>       \
> +     __SMU_DUMMY_MAP(StopGfxOffResidencyLogging),            \
>       __SMU_DUMMY_MAP(SetNumBadMemoryPagesRetired),
>       \
>
>       __SMU_DUMMY_MAP(SetBadMemoryPagesRetiredFlagsPerChannel)
> , \
>       __SMU_DUMMY_MAP(AllowGpo),      \
> diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
> b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
> index 717ba46c8933..1104f8b05b29 100644
> --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
> +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c
> @@ -72,6 +72,12 @@ static const struct smu_feature_bits
> vangogh_dpm_features = {
>       }
>  };
>
> +/*
> + * SMU support new GFXOFF residency log interface since version
> +4.63.62.00,
> + * use this to get live readings of GFXOFF residency  */ #define
> +SUPPORT_LIVE_RESIDENCY_MSG_VERSION 0x043f3e00
> +
>  static struct cmn2asic_msg_mapping
> vangogh_message_map[SMU_MSG_MAX_COUNT] = {
>       MSG_MAP(TestMessage,                    PPSMC_MSG_TestMessage,
>               0),
>       MSG_MAP(GetSmuVersion,                  PPSMC_MSG_GetSmuVersion,
>               0),
> @@ -142,7 +148,9 @@ static struct cmn2asic_msg_mapping
> vangogh_message_map[SMU_MSG_MAX_COUNT] = {
>       MSG_MAP(GetSlowPPTLimit,
> PPSMC_MSG_GetSlowPPTLimit,
>       0),
>       MSG_MAP(GetGfxOffStatus,
> PPSMC_MSG_GetGfxOffStatus,                                            0),
>       MSG_MAP(GetGfxOffEntryCount,
> PPSMC_MSG_GetGfxOffEntryCount,                                        0),
> -     MSG_MAP(LogGfxOffResidency,
> PPSMC_MSG_LogGfxOffResidency,                                 0),
> +     MSG_MAP(StartGfxOffResidencyLogging,
> PPSMC_MSG_StartGfxOffResidencyLogging,        0),
> +     MSG_MAP(GfxOffResidencyLogReadSample,
> PPSMC_MSG_GfxOffResidencyLogReadSample,       0),
> +     MSG_MAP(StopGfxOffResidencyLogging,
> PPSMC_MSG_StopGfxOffResidencyLogging,         0),
>  };
>
>  static struct cmn2asic_mapping
> vangogh_feature_mask_map[SMU_FEATURE_COUNT] = { @@ -2450,19
> +2458,32 @@ static int vangogh_set_power_limit(struct smu_context *smu,
> static u32 vangogh_set_gfxoff_residency(struct smu_context *smu, bool
> start)  {
>       int ret = 0;
> -     u32 residency;
>       struct amdgpu_device *adev = smu->adev;
>
>       if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
>               return 0;
>
> -     ret = smu_cmn_send_smc_msg_with_param(smu,
> SMU_MSG_LogGfxOffResidency,
> -                                           start, &residency);
> -     if (ret)
> -             return ret;
> +     if (smu->smc_fw_version <
> SUPPORT_LIVE_RESIDENCY_MSG_VERSION) {
> +             u32 residency;
> +
> +             ret = smu_cmn_send_smc_msg_with_param(smu,
> SMU_MSG_GfxOffResidencyLogReadSample,
> +                                                     start, &residency);
> +             if (ret)
> +                     return ret;
>
> -     if (!start)
> -             adev->gfx.gfx_off_residency = residency;
> +             if (!start)
> +                     adev->gfx.gfx_off_residency = residency;
> +     } else {
> +             if (start) {
> +                     ret = smu_cmn_send_smc_msg(smu,
> SMU_MSG_StartGfxOffResidencyLogging, NULL);
> +                     if (ret)
> +                             return ret;
> +             } else {
> +                     ret = smu_cmn_send_smc_msg(smu,
> SMU_MSG_StopGfxOffResidencyLogging, NULL);
> +                     if (ret)
> +                             return ret;
> +             }
> +     }
>
>       return ret;
>  }
> @@ -2479,11 +2500,20 @@ static u32 vangogh_set_gfxoff_residency(struct
> smu_context *smu, bool start)
>   */
>  static u32 vangogh_get_gfxoff_residency(struct smu_context *smu, uint32_t
> *residency)  {
> +     int ret = 0;
>       struct amdgpu_device *adev = smu->adev;
>
> -     *residency = adev->gfx.gfx_off_residency;
> +     if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
> +             return 0;
> +
> +     if (smu->smc_fw_version <
> SUPPORT_LIVE_RESIDENCY_MSG_VERSION) {
> +             *residency = adev->gfx.gfx_off_residency;
> +     } else {
> +             ret = smu_cmn_send_smc_msg(smu,
> SMU_MSG_GfxOffResidencyLogReadSample,
> +                                                             residency);
> +     }
>
> -     return 0;
> +     return ret;
>  }
>
>  /**
> --
> 2.43.0