Merge tag 'drm-fixes-2025-05-10' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Weekly drm fixes, bit bigger than last week, but overall amdgpu/xe
  with some ivpu bits and a random few fixes, and dropping the
  ttm_backup struct which wrapped struct file and was recently
  frowned at.

  drm:
   - Fix overflow when generating wedged event

  ttm:
   - Fix documentation
   - Remove struct ttm_backup

  panel:
   - simple: Fix timings for AUO G101EVN010

  amdgpu:
   - DC FP fixes
   - Freesync fix
   - DMUB AUX fixes
   - VCN fix
   - Hibernation fixes
   - HDP fixes

  xe:
   - Prevent PF queue overflow
   - Hold all forcewake during mocs test
   - Remove GSC flush on reset path
   - Fix forcewake put on error path
   - Fix runtime warning when building without svm

  i915:
   - Fix oops on resume after disconnecting DP MST sinks during suspend
   - Fix SPLC num_waiters refcounting

  ivpu:
   - Increase timeouts
   - Fix deadlock in cmdq ioctl
   - Unlock mutices in correct order

  v3d:
   - Avoid memory leak in job handling"

* tag 'drm-fixes-2025-05-10' of https://gitlab.freedesktop.org/drm/kernel: (32 commits)
  drm/i915/dp: Fix determining SST/MST mode during MTP TU state computation
  drm/xe: Add config control for svm flush work
  drm/xe: Release force wake first then runtime power
  drm/xe/gsc: do not flush the GSC worker from the reset path
  drm/xe/tests/mocs: Hold XE_FORCEWAKE_ALL for LNCF regs
  drm/xe: Add page queue multiplier
  drm/amdgpu/hdp7: use memcfg register to post the write for HDP flush
  drm/amdgpu/hdp6: use memcfg register to post the write for HDP flush
  drm/amdgpu/hdp5.2: use memcfg register to post the write for HDP flush
  drm/amdgpu/hdp5: use memcfg register to post the write for HDP flush
  drm/amdgpu/hdp4: use memcfg register to post the write for HDP flush
  drm/amdgpu: fix pm notifier handling
  Revert "drm/amd: Stop evicting resources on APUs in suspend"
  drm/amdgpu/vcn: using separate VCN1_AON_SOC offset
  drm/amd/display: Fix wrong handling for AUX_DEFER case
  drm/amd/display: Copy AUX read reply data whenever length > 0
  drm/amd/display: Remove incorrect checking in dmub aux handler
  drm/amd/display: Fix the checking condition in dmub aux handling
  drm/amd/display: Shift DMUB AUX reply command if necessary
  drm/amd/display: Call FP Protect Before Mode Programming/Mode Support
  ...

Author: torvalds

drivers/accel/ivpu/ivpu_hw.c

@@ -119,7 +119,7 @@ static void timeouts_init(struct ivpu_device *vdev)
 		else
 			vdev->timeout.autosuspend = 100;
 		vdev->timeout.d0i3_entry_msg = 5;
-		vdev->timeout.state_dump_msg = 10;
+		vdev->timeout.state_dump_msg = 100;
 	}
 }
 

drivers/accel/ivpu/ivpu_job.c

@@ -681,8 +681,8 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority, u32 cmdq_id)
 err_erase_xa:
 	xa_erase(&vdev->submitted_jobs_xa, job->job_id);
 err_unlock:
-	mutex_unlock(&vdev->submitted_jobs_lock);
 	mutex_unlock(&file_priv->lock);
+	mutex_unlock(&vdev->submitted_jobs_lock);
 	ivpu_rpm_put(vdev);
 	return ret;
 }
@@ -874,15 +874,21 @@ int ivpu_cmdq_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *
 int ivpu_cmdq_create_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 {
 	struct ivpu_file_priv *file_priv = file->driver_priv;
+	struct ivpu_device *vdev = file_priv->vdev;
 	struct drm_ivpu_cmdq_create *args = data;
 	struct ivpu_cmdq *cmdq;
+	int ret;
 
-	if (!ivpu_is_capable(file_priv->vdev, DRM_IVPU_CAP_MANAGE_CMDQ))
+	if (!ivpu_is_capable(vdev, DRM_IVPU_CAP_MANAGE_CMDQ))
 		return -ENODEV;
 
 	if (args->priority > DRM_IVPU_JOB_PRIORITY_REALTIME)
 		return -EINVAL;
 
+	ret = ivpu_rpm_get(vdev);
+	if (ret < 0)
+		return ret;
+
 	mutex_lock(&file_priv->lock);
 
 	cmdq = ivpu_cmdq_create(file_priv, ivpu_job_to_jsm_priority(args->priority), false);
@@ -891,6 +897,8 @@ int ivpu_cmdq_create_ioctl(struct drm_device *dev, void *data, struct drm_file *
 
 	mutex_unlock(&file_priv->lock);
 
+	ivpu_rpm_put(vdev);
+
 	return cmdq ? 0 : -ENOMEM;
 }
 
@@ -900,28 +908,35 @@ int ivpu_cmdq_destroy_ioctl(struct drm_device *dev, void *data, struct drm_file
 	struct ivpu_device *vdev = file_priv->vdev;
 	struct drm_ivpu_cmdq_destroy *args = data;
 	struct ivpu_cmdq *cmdq;
-	u32 cmdq_id;
+	u32 cmdq_id = 0;
 	int ret;
 
 	if (!ivpu_is_capable(vdev, DRM_IVPU_CAP_MANAGE_CMDQ))
 		return -ENODEV;
 
+	ret = ivpu_rpm_get(vdev);
+	if (ret < 0)
+		return ret;
+
 	mutex_lock(&file_priv->lock);
 
 	cmdq = xa_load(&file_priv->cmdq_xa, args->cmdq_id);
 	if (!cmdq || cmdq->is_legacy) {
 		ret = -ENOENT;
-		goto err_unlock;
+	} else {
+		cmdq_id = cmdq->id;
+		ivpu_cmdq_destroy(file_priv, cmdq);
+		ret = 0;
 	}
 
-	cmdq_id = cmdq->id;
-	ivpu_cmdq_destroy(file_priv, cmdq);
 	mutex_unlock(&file_priv->lock);
-	ivpu_cmdq_abort_all_jobs(vdev, file_priv->ctx.id, cmdq_id);
-	return 0;
 
-err_unlock:
-	mutex_unlock(&file_priv->lock);
+	/* Abort any pending jobs only if cmdq was destroyed */
+	if (!ret)
+		ivpu_cmdq_abort_all_jobs(vdev, file_priv->ctx.id, cmdq_id);
+
+	ivpu_rpm_put(vdev);
+
 	return ret;
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -1614,11 +1614,9 @@ static inline void amdgpu_acpi_get_backlight_caps(struct amdgpu_dm_backlight_cap
 #if defined(CONFIG_ACPI) && defined(CONFIG_SUSPEND)
 bool amdgpu_acpi_is_s3_active(struct amdgpu_device *adev);
 bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev);
-void amdgpu_choose_low_power_state(struct amdgpu_device *adev);
 #else
 static inline bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev) { return false; }
 static inline bool amdgpu_acpi_is_s3_active(struct amdgpu_device *adev) { return false; }
-static inline void amdgpu_choose_low_power_state(struct amdgpu_device *adev) { }
 #endif
 
 void amdgpu_register_gpu_instance(struct amdgpu_device *adev);

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

@@ -1533,22 +1533,4 @@ bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev)
 #endif /* CONFIG_AMD_PMC */
 }
 
-/**
- * amdgpu_choose_low_power_state
- *
- * @adev: amdgpu_device_pointer
- *
- * Choose the target low power state for the GPU
- */
-void amdgpu_choose_low_power_state(struct amdgpu_device *adev)
-{
-	if (adev->in_runpm)
-		return;
-
-	if (amdgpu_acpi_is_s0ix_active(adev))
-		adev->in_s0ix = true;
-	else if (amdgpu_acpi_is_s3_active(adev))
-		adev->in_s3 = true;
-}
-
 #endif /* CONFIG_SUSPEND */

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

@@ -4907,28 +4907,20 @@ static int amdgpu_device_evict_resources(struct amdgpu_device *adev)
  * @data: data
  *
  * This function is called when the system is about to suspend or hibernate.
- * It is used to evict resources from the device before the system goes to
- * sleep while there is still access to swap.
+ * It is used to set the appropriate flags so that eviction can be optimized
+ * in the pm prepare callback.
  */
 static int amdgpu_device_pm_notifier(struct notifier_block *nb, unsigned long mode,
 				     void *data)
 {
 	struct amdgpu_device *adev = container_of(nb, struct amdgpu_device, pm_nb);
-	int r;
 
 	switch (mode) {
 	case PM_HIBERNATION_PREPARE:
 		adev->in_s4 = true;
-		fallthrough;
-	case PM_SUSPEND_PREPARE:
-		r = amdgpu_device_evict_resources(adev);
-		/*
-		 * This is considered non-fatal at this time because
-		 * amdgpu_device_prepare() will also fatally evict resources.
-		 * See https://gitlab.freedesktop.org/drm/amd/-/issues/3781
-		 */
-		if (r)
-			drm_warn(adev_to_drm(adev), "Failed to evict resources, freeze active processes if problems occur: %d\n", r);
+		break;
+	case PM_POST_HIBERNATION:
+		adev->in_s4 = false;
 		break;
 	}
 
@@ -4949,15 +4941,13 @@ int amdgpu_device_prepare(struct drm_device *dev)
 	struct amdgpu_device *adev = drm_to_adev(dev);
 	int i, r;
 
-	amdgpu_choose_low_power_state(adev);
-
 	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
 	/* Evict the majority of BOs before starting suspend sequence */
 	r = amdgpu_device_evict_resources(adev);
 	if (r)
-		goto unprepare;
+		return r;
 
 	flush_delayed_work(&adev->gfx.gfx_off_delay_work);
 
@@ -4968,15 +4958,10 @@ int amdgpu_device_prepare(struct drm_device *dev)
 			continue;
 		r = adev->ip_blocks[i].version->funcs->prepare_suspend(&adev->ip_blocks[i]);
 		if (r)
-			goto unprepare;
+			return r;
 	}
 
 	return 0;
-
-unprepare:
-	adev->in_s0ix = adev->in_s3 = adev->in_s4 = false;
-
-	return r;
 }
 
 /**

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

@@ -2615,13 +2615,8 @@ static int amdgpu_pmops_freeze(struct device *dev)
 static int amdgpu_pmops_thaw(struct device *dev)
 {
 	struct drm_device *drm_dev = dev_get_drvdata(dev);
-	struct amdgpu_device *adev = drm_to_adev(drm_dev);
-	int r;
-
-	r = amdgpu_device_resume(drm_dev, true);
-	adev->in_s4 = false;
 
-	return r;
+	return amdgpu_device_resume(drm_dev, true);
 }
 
 static int amdgpu_pmops_poweroff(struct device *dev)
@@ -2634,9 +2629,6 @@ static int amdgpu_pmops_poweroff(struct device *dev)
 static int amdgpu_pmops_restore(struct device *dev)
 {
 	struct drm_device *drm_dev = dev_get_drvdata(dev);
-	struct amdgpu_device *adev = drm_to_adev(drm_dev);
-
-	adev->in_s4 = false;
 
 	return amdgpu_device_resume(drm_dev, true);
 }

drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h

@@ -66,7 +66,6 @@
 #define VCN_ENC_CMD_REG_WAIT		0x0000000c
 
 #define VCN_AON_SOC_ADDRESS_2_0 	0x1f800
-#define VCN1_AON_SOC_ADDRESS_3_0 	0x48000
 #define VCN_VID_IP_ADDRESS_2_0		0x0
 #define VCN_AON_IP_ADDRESS_2_0		0x30000
 

drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c

@@ -41,7 +41,12 @@ static void hdp_v4_0_flush_hdp(struct amdgpu_device *adev,
 {
 	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+		/* We just need to read back a register to post the write.
+		 * Reading back the remapped register causes problems on
+		 * some platforms so just read back the memory size register.
+		 */
+		if (adev->nbio.funcs->get_memsize)
+			adev->nbio.funcs->get_memsize(adev);
 	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
 	}

drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c

@@ -32,7 +32,12 @@ static void hdp_v5_0_flush_hdp(struct amdgpu_device *adev,
 {
 	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
-		RREG32((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+		/* We just need to read back a register to post the write.
+		 * Reading back the remapped register causes problems on
+		 * some platforms so just read back the memory size register.
+		 */
+		if (adev->nbio.funcs->get_memsize)
+			adev->nbio.funcs->get_memsize(adev);
 	} else {
 		amdgpu_ring_emit_wreg(ring, (adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2, 0);
 	}

drivers/gpu/drm/amd/amdgpu/hdp_v5_2.c

@@ -33,7 +33,17 @@ static void hdp_v5_2_flush_hdp(struct amdgpu_device *adev,
 	if (!ring || !ring->funcs->emit_wreg) {
 		WREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2,
 			0);
-		RREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+		if (amdgpu_sriov_vf(adev)) {
+			/* this is fine because SR_IOV doesn't remap the register */
+			RREG32_NO_KIQ((adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2);
+		} else {
+			/* We just need to read back a register to post the write.
+			 * Reading back the remapped register causes problems on
+			 * some platforms so just read back the memory size register.
+			 */
+			if (adev->nbio.funcs->get_memsize)
+				adev->nbio.funcs->get_memsize(adev);
+		}
 	} else {
 		amdgpu_ring_emit_wreg(ring,
 			(adev->rmmio_remap.reg_offset + KFD_MMIO_REMAP_HDP_MEM_FLUSH_CNTL) >> 2,