[PATCH 23/66] drm/amdgpu/amdgpu_smu: convert to IP version checking
Lazar, Lijo
lijo.lazar at amd.com
Wed Sep 22 07:32:40 UTC 2021
On 9/21/2021 11:36 PM, Alex Deucher wrote:
> Use IP versions rather than asic_type to differentiate
> IP version specific features.
>
> v2: rebase
>
> Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
> ---
> drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c | 94 +++++++++++++----------
> 1 file changed, 55 insertions(+), 39 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
> index 04863a797115..5f372d353d9d 100644
> --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
> +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c
> @@ -455,7 +455,8 @@ static int smu_get_power_num_states(void *handle,
>
> bool is_support_sw_smu(struct amdgpu_device *adev)
> {
> - if (adev->asic_type >= CHIP_ARCTURUS)
> + if ((adev->asic_type >= CHIP_ARCTURUS) ||
> + (adev->ip_versions[MP1_HWIP] >= IP_VERSION(11, 0, 0)))
> return true;
>
> return false;
> @@ -575,43 +576,47 @@ static int smu_set_funcs(struct amdgpu_device *adev)
> if (adev->pm.pp_feature & PP_OVERDRIVE_MASK)
> smu->od_enabled = true;
>
> - switch (adev->asic_type) {
> - case CHIP_NAVI10:
> - case CHIP_NAVI14:
> - case CHIP_NAVI12:
> + switch (adev->ip_versions[MP1_HWIP]) {
> + case IP_VERSION(11, 0, 0):
> + case IP_VERSION(11, 0, 5):
> + case IP_VERSION(11, 0, 9):
> navi10_set_ppt_funcs(smu);
> break;
> - case CHIP_ARCTURUS:
> - adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
> - arcturus_set_ppt_funcs(smu);
> - /* OD is not supported on Arcturus */
> - smu->od_enabled =false;
> - break;
> - case CHIP_SIENNA_CICHLID:
> - case CHIP_NAVY_FLOUNDER:
> - case CHIP_DIMGREY_CAVEFISH:
> - case CHIP_BEIGE_GOBY:
> + case IP_VERSION(11, 0, 7):
> + case IP_VERSION(11, 0, 11):
> + case IP_VERSION(11, 0, 12):
> + case IP_VERSION(11, 0, 13):
> sienna_cichlid_set_ppt_funcs(smu);
> break;
> - case CHIP_ALDEBARAN:
> - aldebaran_set_ppt_funcs(smu);
> - /* Enable pp_od_clk_voltage node */
> - smu->od_enabled = true;
> - break;
> - case CHIP_RENOIR:
> + case IP_VERSION(12, 0, 0):
> renoir_set_ppt_funcs(smu);
> break;
> - case CHIP_VANGOGH:
> + case IP_VERSION(11, 5, 0):
> vangogh_set_ppt_funcs(smu);
> break;
> - case CHIP_YELLOW_CARP:
> + case IP_VERSION(13, 0, 1):
> yellow_carp_set_ppt_funcs(smu);
> break;
> - case CHIP_CYAN_SKILLFISH:
> + case IP_VERSION(11, 0, 8):
> cyan_skillfish_set_ppt_funcs(smu);
> break;
> default:
> - return -EINVAL;
> + switch (adev->asic_type) {
> + case CHIP_ARCTURUS:
> + adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
> + arcturus_set_ppt_funcs(smu);
> + /* OD is not supported on Arcturus */
> + smu->od_enabled =false;
> + break;
> + case CHIP_ALDEBARAN:
> + aldebaran_set_ppt_funcs(smu);
> + /* Enable pp_od_clk_voltage node */
> + smu->od_enabled = true;
> + break;
> + default:
> + return -EINVAL;
> + }
> + break;
> }
>
> return 0;
> @@ -694,7 +699,7 @@ static int smu_late_init(void *handle)
> return ret;
> }
>
> - if (adev->asic_type == CHIP_YELLOW_CARP)
> + if (adev->ip_versions[MP1_HWIP] == IP_VERSION(13, 0, 1))
> return 0;
>
> if (!amdgpu_sriov_vf(adev) || smu->od_enabled) {
> @@ -1140,8 +1145,10 @@ static int smu_smc_hw_setup(struct smu_context *smu)
> if (adev->in_suspend && smu_is_dpm_running(smu)) {
> dev_info(adev->dev, "dpm has been enabled\n");
> /* this is needed specifically */
> - if ((adev->asic_type >= CHIP_SIENNA_CICHLID) &&
> - (adev->asic_type <= CHIP_DIMGREY_CAVEFISH))
> + if ((adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 7)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 11)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 5, 0)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 12)))
switch...case looks better here
> ret = smu_system_features_control(smu, true);
> return ret;
> }
> @@ -1284,7 +1291,7 @@ static int smu_start_smc_engine(struct smu_context *smu)
> int ret = 0;
>
> if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
> - if (adev->asic_type < CHIP_NAVI10) {
> + if (adev->ip_versions[MP1_HWIP] < IP_VERSION(11, 0, 0)) {
> if (smu->ppt_funcs->load_microcode) {
> ret = smu->ppt_funcs->load_microcode(smu);
> if (ret)
> @@ -1403,8 +1410,14 @@ static int smu_disable_dpms(struct smu_context *smu)
> * properly.
> */
> if (smu->uploading_custom_pp_table &&
> - (adev->asic_type >= CHIP_NAVI10) &&
> - (adev->asic_type <= CHIP_BEIGE_GOBY))
> + ((adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 0)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 5)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 9)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 7)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 11)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 5, 0)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 12)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 13))))
Same - switch...case
> return smu_disable_all_features_with_exception(smu,
> true,
> SMU_FEATURE_COUNT);
> @@ -1413,9 +1426,11 @@ static int smu_disable_dpms(struct smu_context *smu)
> * For Sienna_Cichlid, PMFW will handle the features disablement properly
> * on BACO in. Driver involvement is unnecessary.
> */
> - if (((adev->asic_type == CHIP_SIENNA_CICHLID) ||
> - ((adev->asic_type >= CHIP_NAVI10) && (adev->asic_type <= CHIP_NAVI12))) &&
> - use_baco)
> + if (((adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 7)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 0)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 5)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 9))) &&
> + use_baco)
Here as well.
Thanks,
Lijo
> return smu_disable_all_features_with_exception(smu,
> true,
> SMU_FEATURE_BACO_BIT);
> @@ -1436,7 +1451,7 @@ static int smu_disable_dpms(struct smu_context *smu)
> dev_err(adev->dev, "Failed to disable smu features.\n");
> }
>
> - if (adev->asic_type >= CHIP_NAVI10 &&
> + if (adev->ip_versions[MP1_HWIP] >= IP_VERSION(11, 0, 0) &&
> adev->gfx.rlc.funcs->stop)
> adev->gfx.rlc.funcs->stop(adev);
>
> @@ -2229,6 +2244,7 @@ int smu_get_power_limit(void *handle,
> enum pp_power_type pp_power_type)
> {
> struct smu_context *smu = handle;
> + struct amdgpu_device *adev = smu->adev;
> enum smu_ppt_limit_level limit_level;
> uint32_t limit_type;
> int ret = 0;
> @@ -2273,10 +2289,10 @@ int smu_get_power_limit(void *handle,
> switch (limit_level) {
> case SMU_PPT_LIMIT_CURRENT:
> if ((smu->adev->asic_type == CHIP_ALDEBARAN) ||
> - (smu->adev->asic_type == CHIP_SIENNA_CICHLID) ||
> - (smu->adev->asic_type == CHIP_NAVY_FLOUNDER) ||
> - (smu->adev->asic_type == CHIP_DIMGREY_CAVEFISH) ||
> - (smu->adev->asic_type == CHIP_BEIGE_GOBY))
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 7)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 11)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 12)) ||
> + (adev->ip_versions[MP1_HWIP] == IP_VERSION(11, 0, 13)))
> ret = smu_get_asic_power_limits(smu,
> &smu->current_power_limit,
> NULL,
>
More information about the amd-gfx
mailing list