Merge tag 'amd-drm-fixes-6.1-2022-11-02' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.1-2022-11-02:

amdgpu:
- DCN 3.1.4 fixes
- DCN 3.2.x fixes
- GC 11.x fixes
- Virtual display fix
- Fail suspend if resources can't be evicted
- SR-IOV fix
- Display PSR fix

amdkfd:
- Fix possible NULL pointer deref
- GC 11.x trap handler fix

Signed-off-by: Dave Airlie <[email protected]>
From: Alex Deucher <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

Author: airlied

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

@@ -706,6 +706,13 @@ int amdgpu_amdkfd_submit_ib(struct amdgpu_device *adev,
 
 void amdgpu_amdkfd_set_compute_idle(struct amdgpu_device *adev, bool idle)
 {
+	/* Temporary workaround to fix issues observed in some
+	 * compute applications when GFXOFF is enabled on GFX11.
+	 */
+	if (IP_VERSION_MAJ(adev->ip_versions[GC_HWIP][0]) == 11) {
+		pr_debug("GFXOFF is %s\n", idle ? "enabled" : "disabled");
+		amdgpu_gfx_off_ctrl(adev, idle);
+	}
 	amdgpu_dpm_switch_power_profile(adev,
 					PP_SMC_POWER_PROFILE_COMPUTE,
 					!idle);

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -4060,15 +4060,18 @@ void amdgpu_device_fini_sw(struct amdgpu_device *adev)
  * at suspend time.
  *
  */
-static void amdgpu_device_evict_resources(struct amdgpu_device *adev)
+static int amdgpu_device_evict_resources(struct amdgpu_device *adev)
 {
+	int ret;
+
 	/* No need to evict vram on APUs for suspend to ram or s2idle */
 	if ((adev->in_s3 || adev->in_s0ix) && (adev->flags & AMD_IS_APU))
-		return;
+		return 0;
 
-	if (amdgpu_ttm_evict_resources(adev, TTM_PL_VRAM))
+	ret = amdgpu_ttm_evict_resources(adev, TTM_PL_VRAM);
+	if (ret)
 		DRM_WARN("evicting device resources failed\n");
-
+	return ret;
 }
 
 /*
@@ -4118,7 +4121,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon)
 	if (!adev->in_s0ix)
 		amdgpu_amdkfd_suspend(adev, adev->in_runpm);
 
-	amdgpu_device_evict_resources(adev);
+	r = amdgpu_device_evict_resources(adev);
+	if (r)
+		return r;
 
 	amdgpu_fence_driver_hw_fini(adev);
 

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -2201,7 +2201,8 @@ amdgpu_pci_remove(struct pci_dev *pdev)
 		pm_runtime_forbid(dev->dev);
 	}
 
-	if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 2)) {
+	if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 2) &&
+	    !amdgpu_sriov_vf(adev)) {
 		bool need_to_reset_gpu = false;
 
 		if (adev->gmc.xgmi.num_physical_nodes > 1) {

drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c

@@ -337,12 +337,14 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		fw_info->feature = adev->psp.cap_feature_version;
 		break;
 	case AMDGPU_INFO_FW_MES_KIQ:
-		fw_info->ver = adev->mes.ucode_fw_version[0];
-		fw_info->feature = 0;
+		fw_info->ver = adev->mes.kiq_version & AMDGPU_MES_VERSION_MASK;
+		fw_info->feature = (adev->mes.kiq_version & AMDGPU_MES_FEAT_VERSION_MASK)
+					>> AMDGPU_MES_FEAT_VERSION_SHIFT;
 		break;
 	case AMDGPU_INFO_FW_MES:
-		fw_info->ver = adev->mes.ucode_fw_version[1];
-		fw_info->feature = 0;
+		fw_info->ver = adev->mes.sched_version & AMDGPU_MES_VERSION_MASK;
+		fw_info->feature = (adev->mes.sched_version & AMDGPU_MES_FEAT_VERSION_MASK)
+					>> AMDGPU_MES_FEAT_VERSION_SHIFT;
 		break;
 	case AMDGPU_INFO_FW_IMU:
 		fw_info->ver = adev->gfx.imu_fw_version;

drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c

@@ -500,6 +500,8 @@ static int amdgpu_vkms_sw_init(void *handle)
 
 	adev_to_drm(adev)->mode_config.fb_base = adev->gmc.aper_base;
 
+	adev_to_drm(adev)->mode_config.fb_modifiers_not_supported = true;
+
 	r = amdgpu_display_modeset_create_props(adev);
 	if (r)
 		return r;