From: Alex Deucher <alexander.deucher@xxxxxxx> [ Upstream commit 66429300e1bd9cdfbe96cfc475e4964db2a36921 ] The call to pm_runtime_get_sync increments the counter even in case of failure, leading to incorrect ref count. In case of failure, decrement the ref count before returning. Reviewed-by: Evan Quan <evan.quan@xxxxxxx> Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> --- drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 220 ++++++++++++++++++------- 1 file changed, 165 insertions(+), 55 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 96b8feb77b15d..e08b0033ca465 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -167,8 +167,10 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { if (adev->smu.ppt_funcs->get_current_power_state) @@ -212,8 +214,10 @@ static ssize_t amdgpu_set_dpm_state(struct device *dev, return -EINVAL; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { mutex_lock(&adev->pm.mutex); @@ -307,8 +311,10 @@ static ssize_t amdgpu_get_dpm_forced_performance_level(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) level = smu_get_performance_level(&adev->smu); @@ -369,8 +375,10 @@ static ssize_t amdgpu_set_dpm_forced_performance_level(struct device *dev, } ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) current_level = smu_get_performance_level(&adev->smu); @@ -446,8 +454,10 @@ static ssize_t amdgpu_get_pp_num_states(struct device *dev, int i, buf_len, ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { ret = smu_get_power_num_states(&adev->smu, &data); @@ -488,8 +498,10 @@ static ssize_t amdgpu_get_pp_cur_state(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { pm = smu_get_current_power_state(smu); @@ -564,8 +576,10 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, state = data.states[idx]; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } /* only set user selected power states */ if (state != POWER_STATE_TYPE_INTERNAL_BOOT && @@ -605,8 +619,10 @@ static ssize_t amdgpu_get_pp_table(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { size = smu_sys_get_pp_table(&adev->smu, (void **)&table); @@ -647,8 +663,10 @@ static ssize_t amdgpu_set_pp_table(struct device *dev, return -EINVAL; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { ret = smu_sys_set_pp_table(&adev->smu, (void *)buf, count); @@ -787,8 +805,10 @@ static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev, } ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { ret = smu_od_edit_dpm_table(&adev->smu, type, @@ -844,8 +864,10 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { size = smu_print_clk_levels(&adev->smu, SMU_OD_SCLK, buf); @@ -902,8 +924,10 @@ static ssize_t amdgpu_set_pp_feature_status(struct device *dev, pr_debug("featuremask = 0x%llx\n", featuremask); ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { ret = smu_sys_set_pp_feature_mask(&adev->smu, featuremask); @@ -939,8 +963,10 @@ static ssize_t amdgpu_get_pp_feature_status(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_sys_get_pp_feature_mask(&adev->smu, buf); @@ -998,8 +1024,10 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_SCLK, buf); @@ -1068,8 +1096,10 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_SCLK, mask, true); @@ -1098,8 +1128,10 @@ static ssize_t amdgpu_get_pp_dpm_mclk(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_MCLK, buf); @@ -1132,8 +1164,10 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_MCLK, mask, true); @@ -1162,8 +1196,10 @@ static ssize_t amdgpu_get_pp_dpm_socclk(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_SOCCLK, buf); @@ -1196,8 +1232,10 @@ static ssize_t amdgpu_set_pp_dpm_socclk(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_SOCCLK, mask, true); @@ -1228,8 +1266,10 @@ static ssize_t amdgpu_get_pp_dpm_fclk(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_FCLK, buf); @@ -1262,8 +1302,10 @@ static ssize_t amdgpu_set_pp_dpm_fclk(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_FCLK, mask, true); @@ -1294,8 +1336,10 @@ static ssize_t amdgpu_get_pp_dpm_dcefclk(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_DCEFCLK, buf); @@ -1328,8 +1372,10 @@ static ssize_t amdgpu_set_pp_dpm_dcefclk(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_DCEFCLK, mask, true); @@ -1360,8 +1406,10 @@ static ssize_t amdgpu_get_pp_dpm_pcie(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_print_clk_levels(&adev->smu, SMU_PCIE, buf); @@ -1394,8 +1442,10 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev, return ret; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_force_clk_levels(&adev->smu, SMU_PCIE, mask, true); @@ -1426,8 +1476,10 @@ static ssize_t amdgpu_get_pp_sclk_od(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) value = smu_get_od_percentage(&(adev->smu), SMU_OD_SCLK); @@ -1459,8 +1511,10 @@ static ssize_t amdgpu_set_pp_sclk_od(struct device *dev, return -EINVAL; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { value = smu_set_od_percentage(&(adev->smu), SMU_OD_SCLK, (uint32_t)value); @@ -1495,8 +1549,10 @@ static ssize_t amdgpu_get_pp_mclk_od(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) value = smu_get_od_percentage(&(adev->smu), SMU_OD_MCLK); @@ -1528,8 +1584,10 @@ static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, return -EINVAL; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { value = smu_set_od_percentage(&(adev->smu), SMU_OD_MCLK, (uint32_t)value); @@ -1584,8 +1642,10 @@ static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) size = smu_get_power_profile_mode(&adev->smu, buf); @@ -1647,8 +1707,10 @@ static ssize_t amdgpu_set_pp_power_profile_mode(struct device *dev, parameter[parameter_size] = profile_mode; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } if (is_support_sw_smu(adev)) ret = smu_set_power_profile_mode(&adev->smu, parameter, parameter_size, true); @@ -1684,8 +1746,10 @@ static ssize_t amdgpu_get_busy_percent(struct device *dev, return 0; r = pm_runtime_get_sync(ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(ddev->dev); return r; + } /* read the IP busy sensor */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_LOAD, @@ -1720,8 +1784,10 @@ static ssize_t amdgpu_get_memory_busy_percent(struct device *dev, return 0; r = pm_runtime_get_sync(ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(ddev->dev); return r; + } /* read the IP busy sensor */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MEM_LOAD, @@ -1761,8 +1827,10 @@ static ssize_t amdgpu_get_pcie_bw(struct device *dev, return 0; ret = pm_runtime_get_sync(ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(ddev->dev); return ret; + } amdgpu_asic_get_pcie_usage(adev, &count0, &count1); @@ -1863,8 +1931,10 @@ static ssize_t amdgpu_hwmon_show_temp(struct device *dev, return -EINVAL; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } switch (channel) { case PP_TEMP_JUNCTION: @@ -1991,8 +2061,10 @@ static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev, int ret; ret = pm_runtime_get_sync(adev->ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { pwm_mode = smu_get_fan_control_mode(&adev->smu); @@ -2026,8 +2098,10 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, return err; ret = pm_runtime_get_sync(adev->ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { smu_set_fan_control_mode(&adev->smu, value); @@ -2071,8 +2145,10 @@ static ssize_t amdgpu_hwmon_set_pwm1(struct device *dev, u32 pwm_mode; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) pwm_mode = smu_get_fan_control_mode(&adev->smu); @@ -2120,8 +2196,10 @@ static ssize_t amdgpu_hwmon_get_pwm1(struct device *dev, u32 speed = 0; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) err = smu_get_fan_speed_percent(&adev->smu, &speed); @@ -2150,8 +2228,10 @@ static ssize_t amdgpu_hwmon_get_fan1_input(struct device *dev, u32 speed = 0; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) err = smu_get_fan_speed_rpm(&adev->smu, &speed); @@ -2179,8 +2259,10 @@ static ssize_t amdgpu_hwmon_get_fan1_min(struct device *dev, int r; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MIN_FAN_RPM, (void *)&min_rpm, &size); @@ -2204,8 +2286,10 @@ static ssize_t amdgpu_hwmon_get_fan1_max(struct device *dev, int r; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_MAX_FAN_RPM, (void *)&max_rpm, &size); @@ -2228,8 +2312,10 @@ static ssize_t amdgpu_hwmon_get_fan1_target(struct device *dev, u32 rpm = 0; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) err = smu_get_fan_speed_rpm(&adev->smu, &rpm); @@ -2257,8 +2343,10 @@ static ssize_t amdgpu_hwmon_set_fan1_target(struct device *dev, u32 pwm_mode; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) pwm_mode = smu_get_fan_control_mode(&adev->smu); @@ -2303,8 +2391,10 @@ static ssize_t amdgpu_hwmon_get_fan1_enable(struct device *dev, int ret; ret = pm_runtime_get_sync(adev->ddev->dev); - if (ret < 0) + if (ret < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return ret; + } if (is_support_sw_smu(adev)) { pwm_mode = smu_get_fan_control_mode(&adev->smu); @@ -2346,8 +2436,10 @@ static ssize_t amdgpu_hwmon_set_fan1_enable(struct device *dev, return -EINVAL; err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) { smu_set_fan_control_mode(&adev->smu, pwm_mode); @@ -2375,8 +2467,10 @@ static ssize_t amdgpu_hwmon_show_vddgfx(struct device *dev, int r, size = sizeof(vddgfx); r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDGFX, @@ -2411,8 +2505,10 @@ static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev, return -EINVAL; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, @@ -2444,8 +2540,10 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev, unsigned uw; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } /* get the voltage */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, @@ -2480,8 +2578,10 @@ static ssize_t amdgpu_hwmon_show_power_cap_max(struct device *dev, int r; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } if (is_support_sw_smu(adev)) { smu_get_power_limit(&adev->smu, &limit, true, true); @@ -2509,8 +2609,10 @@ static ssize_t amdgpu_hwmon_show_power_cap(struct device *dev, int r; r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } if (is_support_sw_smu(adev)) { smu_get_power_limit(&adev->smu, &limit, false, true); @@ -2549,8 +2651,10 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev, err = pm_runtime_get_sync(adev->ddev->dev); - if (err < 0) + if (err < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return err; + } if (is_support_sw_smu(adev)) err = smu_set_power_limit(&adev->smu, value); @@ -2577,8 +2681,10 @@ static ssize_t amdgpu_hwmon_show_sclk(struct device *dev, int r, size = sizeof(sclk); r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } /* get the sclk */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_SCLK, @@ -2609,8 +2715,10 @@ static ssize_t amdgpu_hwmon_show_mclk(struct device *dev, int r, size = sizeof(mclk); r = pm_runtime_get_sync(adev->ddev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(adev->ddev->dev); return r; + } /* get the sclk */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_MCLK, @@ -3639,8 +3747,10 @@ static int amdgpu_debugfs_pm_info(struct seq_file *m, void *data) int r; r = pm_runtime_get_sync(dev->dev); - if (r < 0) + if (r < 0) { + pm_runtime_put_autosuspend(dev->dev); return r; + } amdgpu_device_ip_get_clockgating_state(adev, &flags); seq_printf(m, "Clock Gating Flags Mask: 0x%x\n", flags); -- 2.25.1 _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel