Merge tag 'drm-fixes-2020-05-15' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "As mentioned last week an i915 PR came in late, but I left it, so the
  i915 bits of this cover 2 weeks, which is why it's likely a bit larger
  than usual.

  Otherwise it's mostly amdgpu fixes, one tegra fix, one meson fix.

  i915:
   - Handle idling during i915_gem_evict_something busy loops (Chris)
   - Mark current submissions with a weak-dependency (Chris)
   - Propagate error from completed fences (Chris)
   - Fixes on execlist to avoid GPU hang situation (Chris)
   - Fixes couple deadlocks (Chris)
   - Timeslice preemption fixes (Chris)
   - Fix Display Port interrupt handling on Tiger Lake (Imre)
   - Reduce debug noise around Frame Buffer Compression (Peter)
   - Fix logic around IPC W/a for Coffee Lake and Kaby Lake (Sultan)
   - Avoid dereferencing a dead context (Chris)

  tegra:
   - tegra120/4 smmu fixes

  amdgpu:
   - Clockgating fixes
   - Fix fbdev with scatter/gather display
   - S4 fix for navi
   - Soft recovery for gfx10
   - Freesync fixes
   - Atomic check cursor fix
   - Add a gfxoff quirk
   - MST fix

  amdkfd:
   - Fix GEM reference counting

  meson:
   - error code propogation fix"

* tag 'drm-fixes-2020-05-15' of git://anongit.freedesktop.org/drm/drm: (29 commits)
  drm/i915: Handle idling during i915_gem_evict_something busy loops
  drm/meson: pm resume add return errno branch
  drm/amd/amdgpu: Update update_config() logic
  drm/amd/amdgpu: add raven1 part to the gfxoff quirk list
  drm/i915: Mark concurrent submissions with a weak-dependency
  drm/i915: Propagate error from completed fences
  drm/i915/gvt: Fix kernel oops for 3-level ppgtt guest
  drm/i915/gvt: Init DPLL/DDI vreg for virtual display instead of inheritance.
  drm/amd/display: add basic atomic check for cursor plane
  drm/amd/display: Fix vblank and pageflip event handling for FreeSync
  drm/amdgpu: implement soft_recovery for gfx10
  drm/amdgpu: enable hibernate support on Navi1X
  drm/amdgpu: Use GEM obj reference for KFD BOs
  drm/amdgpu: force fbdev into vram
  drm/amd/powerplay: perform PG ungate prior to CG ungate
  drm/amdgpu: drop unnecessary cancel_delayed_work_sync on PG ungate
  drm/amdgpu: disable MGCG/MGLS also on gfx CG ungate
  drm/i915/execlists: Track inflight CCID
  drm/i915/execlists: Avoid reusing the same logical CCID
  drm/i915/gem: Remove object_is_locked assertion from unpin_from_display_plane
  ...

Author: torvalds

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

@@ -945,6 +945,7 @@ struct amdgpu_device {
 
 	/* s3/s4 mask */
 	bool                            in_suspend;
+	bool				in_hibernate;
 
 	/* record last mm index being written through WREG32*/
 	unsigned long last_mm_index;

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

@@ -1343,7 +1343,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
 	}
 
 	/* Free the BO*/
-	amdgpu_bo_unref(&mem->bo);
+	drm_gem_object_put_unlocked(&mem->bo->tbo.base);
 	mutex_destroy(&mem->lock);
 	kfree(mem);
 
@@ -1688,7 +1688,8 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
 		| KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE
 		| KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
 
-	(*mem)->bo = amdgpu_bo_ref(bo);
+	drm_gem_object_get(&bo->tbo.base);
+	(*mem)->bo = bo;
 	(*mem)->va = va;
 	(*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
 		AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT;

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

@@ -1181,7 +1181,9 @@ static int amdgpu_pmops_freeze(struct device *dev)
 	struct amdgpu_device *adev = drm_dev->dev_private;
 	int r;
 
+	adev->in_hibernate = true;
 	r = amdgpu_device_suspend(drm_dev, true);
+	adev->in_hibernate = false;
 	if (r)
 		return r;
 	return amdgpu_asic_reset(adev);

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

@@ -133,8 +133,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 	u32 cpp;
 	u64 flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
 			       AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS     |
-			       AMDGPU_GEM_CREATE_VRAM_CLEARED 	     |
-			       AMDGPU_GEM_CREATE_CPU_GTT_USWC;
+			       AMDGPU_GEM_CREATE_VRAM_CLEARED;
 
 	info = drm_get_format_info(adev->ddev, mode_cmd);
 	cpp = info->cpp[0];

drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c

@@ -4273,7 +4273,7 @@ static int gfx_v10_0_update_gfx_clock_gating(struct amdgpu_device *adev,
 		/* ===  CGCG /CGLS for GFX 3D Only === */
 		gfx_v10_0_update_3d_clock_gating(adev, enable);
 		/* ===  MGCG + MGLS === */
-		/* gfx_v10_0_update_medium_grain_clock_gating(adev, enable); */
+		gfx_v10_0_update_medium_grain_clock_gating(adev, enable);
 	}
 
 	if (adev->cg_flags &
@@ -4353,11 +4353,7 @@ static int gfx_v10_0_set_powergating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
-		if (!enable) {
-			amdgpu_gfx_off_ctrl(adev, false);
-			cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
-		} else
-			amdgpu_gfx_off_ctrl(adev, true);
+		amdgpu_gfx_off_ctrl(adev, enable);
 		break;
 	default:
 		break;
@@ -4918,6 +4914,19 @@ static void gfx_v10_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
 							   ref, mask);
 }
 
+static void gfx_v10_0_ring_soft_recovery(struct amdgpu_ring *ring,
+					 unsigned vmid)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t value = 0;
+
+	value = REG_SET_FIELD(value, SQ_CMD, CMD, 0x03);
+	value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01);
+	value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1);
+	value = REG_SET_FIELD(value, SQ_CMD, VM_ID, vmid);
+	WREG32_SOC15(GC, 0, mmSQ_CMD, value);
+}
+
 static void
 gfx_v10_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
 				      uint32_t me, uint32_t pipe,
@@ -5309,6 +5318,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = {
 	.emit_wreg = gfx_v10_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v10_0_ring_emit_reg_wait,
 	.emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait,
+	.soft_recovery = gfx_v10_0_ring_soft_recovery,
 };
 
 static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = {

drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c

@@ -1236,6 +1236,8 @@ static const struct amdgpu_gfxoff_quirk amdgpu_gfxoff_quirk_list[] = {
 	{ 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc8 },
 	/* https://bugzilla.kernel.org/show_bug.cgi?id=207171 */
 	{ 0x1002, 0x15dd, 0x103c, 0x83e7, 0xd3 },
+	/* GFXOFF is unstable on C6 parts with a VBIOS 113-RAVEN-114 */
+	{ 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc6 },
 	{ 0, 0, 0, 0, 0 },
 };
 
@@ -5025,10 +5027,9 @@ static int gfx_v9_0_set_powergating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_RAVEN:
 	case CHIP_RENOIR:
-		if (!enable) {
+		if (!enable)
 			amdgpu_gfx_off_ctrl(adev, false);
-			cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
-		}
+
 		if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) {
 			gfx_v9_0_enable_sck_slow_down_on_power_up(adev, true);
 			gfx_v9_0_enable_sck_slow_down_on_power_down(adev, true);
@@ -5052,12 +5053,7 @@ static int gfx_v9_0_set_powergating_state(void *handle,
 			amdgpu_gfx_off_ctrl(adev, true);
 		break;
 	case CHIP_VEGA12:
-		if (!enable) {
-			amdgpu_gfx_off_ctrl(adev, false);
-			cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
-		} else {
-			amdgpu_gfx_off_ctrl(adev, true);
-		}
+		amdgpu_gfx_off_ctrl(adev, enable);
 		break;
 	default:
 		break;

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -441,7 +441,7 @@ static void dm_vupdate_high_irq(void *interrupt_params)
 
 /**
  * dm_crtc_high_irq() - Handles CRTC interrupt
- * @interrupt_params: ignored
+ * @interrupt_params: used for determining the CRTC instance
  *
  * Handles the CRTC/VSYNC interrupt by notfying DRM's VBLANK
  * event handler.
@@ -455,70 +455,6 @@ static void dm_crtc_high_irq(void *interrupt_params)
 	unsigned long flags;
 
 	acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VBLANK);
-
-	if (acrtc) {
-		acrtc_state = to_dm_crtc_state(acrtc->base.state);
-
-		DRM_DEBUG_VBL("crtc:%d, vupdate-vrr:%d\n",
-			      acrtc->crtc_id,
-			      amdgpu_dm_vrr_active(acrtc_state));
-
-		/* Core vblank handling at start of front-porch is only possible
-		 * in non-vrr mode, as only there vblank timestamping will give
-		 * valid results while done in front-porch. Otherwise defer it
-		 * to dm_vupdate_high_irq after end of front-porch.
-		 */
-		if (!amdgpu_dm_vrr_active(acrtc_state))
-			drm_crtc_handle_vblank(&acrtc->base);
-
-		/* Following stuff must happen at start of vblank, for crc
-		 * computation and below-the-range btr support in vrr mode.
-		 */
-		amdgpu_dm_crtc_handle_crc_irq(&acrtc->base);
-
-		if (acrtc_state->stream && adev->family >= AMDGPU_FAMILY_AI &&
-		    acrtc_state->vrr_params.supported &&
-		    acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE) {
-			spin_lock_irqsave(&adev->ddev->event_lock, flags);
-			mod_freesync_handle_v_update(
-				adev->dm.freesync_module,
-				acrtc_state->stream,
-				&acrtc_state->vrr_params);
-
-			dc_stream_adjust_vmin_vmax(
-				adev->dm.dc,
-				acrtc_state->stream,
-				&acrtc_state->vrr_params.adjust);
-			spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
-		}
-	}
-}
-
-#if defined(CONFIG_DRM_AMD_DC_DCN)
-/**
- * dm_dcn_crtc_high_irq() - Handles VStartup interrupt for DCN generation ASICs
- * @interrupt params - interrupt parameters
- *
- * Notify DRM's vblank event handler at VSTARTUP
- *
- * Unlike DCE hardware, we trigger the handler at VSTARTUP. at which:
- * * We are close enough to VUPDATE - the point of no return for hw
- * * We are in the fixed portion of variable front porch when vrr is enabled
- * * We are before VUPDATE, where double-buffered vrr registers are swapped
- *
- * It is therefore the correct place to signal vblank, send user flip events,
- * and update VRR.
- */
-static void dm_dcn_crtc_high_irq(void *interrupt_params)
-{
-	struct common_irq_params *irq_params = interrupt_params;
-	struct amdgpu_device *adev = irq_params->adev;
-	struct amdgpu_crtc *acrtc;
-	struct dm_crtc_state *acrtc_state;
-	unsigned long flags;
-
-	acrtc = get_crtc_by_otg_inst(adev, irq_params->irq_src - IRQ_TYPE_VBLANK);
-
 	if (!acrtc)
 		return;
 
@@ -528,22 +464,35 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params)
 			 amdgpu_dm_vrr_active(acrtc_state),
 			 acrtc_state->active_planes);
 
+	/**
+	 * Core vblank handling at start of front-porch is only possible
+	 * in non-vrr mode, as only there vblank timestamping will give
+	 * valid results while done in front-porch. Otherwise defer it
+	 * to dm_vupdate_high_irq after end of front-porch.
+	 */
+	if (!amdgpu_dm_vrr_active(acrtc_state))
+		drm_crtc_handle_vblank(&acrtc->base);
+
+	/**
+	 * Following stuff must happen at start of vblank, for crc
+	 * computation and below-the-range btr support in vrr mode.
+	 */
 	amdgpu_dm_crtc_handle_crc_irq(&acrtc->base);
-	drm_crtc_handle_vblank(&acrtc->base);
+
+	/* BTR updates need to happen before VUPDATE on Vega and above. */
+	if (adev->family < AMDGPU_FAMILY_AI)
+		return;
 
 	spin_lock_irqsave(&adev->ddev->event_lock, flags);
 
-	if (acrtc_state->vrr_params.supported &&
+	if (acrtc_state->stream && acrtc_state->vrr_params.supported &&
 	    acrtc_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE) {
-		mod_freesync_handle_v_update(
-		adev->dm.freesync_module,
-		acrtc_state->stream,
-		&acrtc_state->vrr_params);
+		mod_freesync_handle_v_update(adev->dm.freesync_module,
+					     acrtc_state->stream,
+					     &acrtc_state->vrr_params);
 
-		dc_stream_adjust_vmin_vmax(
-			adev->dm.dc,
-			acrtc_state->stream,
-			&acrtc_state->vrr_params.adjust);
+		dc_stream_adjust_vmin_vmax(adev->dm.dc, acrtc_state->stream,
+					   &acrtc_state->vrr_params.adjust);
 	}
 
 	/*
@@ -556,7 +505,8 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params)
 	 * avoid race conditions between flip programming and completion,
 	 * which could cause too early flip completion events.
 	 */
-	if (acrtc->pflip_status == AMDGPU_FLIP_SUBMITTED &&
+	if (adev->family >= AMDGPU_FAMILY_RV &&
+	    acrtc->pflip_status == AMDGPU_FLIP_SUBMITTED &&
 	    acrtc_state->active_planes == 0) {
 		if (acrtc->event) {
 			drm_crtc_send_vblank_event(&acrtc->base, acrtc->event);
@@ -568,7 +518,6 @@ static void dm_dcn_crtc_high_irq(void *interrupt_params)
 
 	spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
 }
-#endif
 
 static int dm_set_clockgating_state(void *handle,
 		  enum amd_clockgating_state state)
@@ -2445,8 +2394,36 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
 		c_irq_params->adev = adev;
 		c_irq_params->irq_src = int_params.irq_source;
 
+		amdgpu_dm_irq_register_interrupt(
+			adev, &int_params, dm_crtc_high_irq, c_irq_params);
+	}
+
+	/* Use VUPDATE_NO_LOCK interrupt on DCN, which seems to correspond to
+	 * the regular VUPDATE interrupt on DCE. We want DC_IRQ_SOURCE_VUPDATEx
+	 * to trigger at end of each vblank, regardless of state of the lock,
+	 * matching DCE behaviour.
+	 */
+	for (i = DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT;
+	     i <= DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT + adev->mode_info.num_crtc - 1;
+	     i++) {
+		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->vupdate_irq);
+
+		if (r) {
+			DRM_ERROR("Failed to add vupdate irq id!\n");
+			return r;
+		}
+
+		int_params.int_context = INTERRUPT_HIGH_IRQ_CONTEXT;
+		int_params.irq_source =
+			dc_interrupt_to_irq_source(dc, i, 0);
+
+		c_irq_params = &adev->dm.vupdate_params[int_params.irq_source - DC_IRQ_SOURCE_VUPDATE1];
+
+		c_irq_params->adev = adev;
+		c_irq_params->irq_src = int_params.irq_source;
+
 		amdgpu_dm_irq_register_interrupt(adev, &int_params,
-				dm_dcn_crtc_high_irq, c_irq_params);
+				dm_vupdate_high_irq, c_irq_params);
 	}
 
 	/* Use GRPH_PFLIP interrupt */
@@ -4453,10 +4430,6 @@ static inline int dm_set_vupdate_irq(struct drm_crtc *crtc, bool enable)
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	int rc;
 
-	/* Do not set vupdate for DCN hardware */
-	if (adev->family > AMDGPU_FAMILY_AI)
-		return 0;
-
 	irq_source = IRQ_TYPE_VUPDATE + acrtc->otg_inst;
 
 	rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY;
@@ -7882,6 +7855,7 @@ static int dm_update_plane_state(struct dc *dc,
 	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
 	struct dm_crtc_state *dm_new_crtc_state, *dm_old_crtc_state;
 	struct dm_plane_state *dm_new_plane_state, *dm_old_plane_state;
+	struct amdgpu_crtc *new_acrtc;
 	bool needs_reset;
 	int ret = 0;
 
@@ -7891,9 +7865,30 @@ static int dm_update_plane_state(struct dc *dc,
 	dm_new_plane_state = to_dm_plane_state(new_plane_state);
 	dm_old_plane_state = to_dm_plane_state(old_plane_state);
 
-	/*TODO Implement atomic check for cursor plane */
-	if (plane->type == DRM_PLANE_TYPE_CURSOR)
+	/*TODO Implement better atomic check for cursor plane */
+	if (plane->type == DRM_PLANE_TYPE_CURSOR) {
+		if (!enable || !new_plane_crtc ||
+			drm_atomic_plane_disabling(plane->state, new_plane_state))
+			return 0;
+
+		new_acrtc = to_amdgpu_crtc(new_plane_crtc);
+
+		if ((new_plane_state->crtc_w > new_acrtc->max_cursor_width) ||
+			(new_plane_state->crtc_h > new_acrtc->max_cursor_height)) {
+			DRM_DEBUG_ATOMIC("Bad cursor size %d x %d\n",
+							 new_plane_state->crtc_w, new_plane_state->crtc_h);
+			return -EINVAL;
+		}
+
+		if (new_plane_state->crtc_x <= -new_acrtc->max_cursor_width ||
+			new_plane_state->crtc_y <= -new_acrtc->max_cursor_height) {
+			DRM_DEBUG_ATOMIC("Bad cursor position %d, %d\n",
+							 new_plane_state->crtc_x, new_plane_state->crtc_y);
+			return -EINVAL;
+		}
+
 		return 0;
+	}
 
 	needs_reset = should_reset_plane(state, plane, old_plane_state,
 					 new_plane_state);

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c

@@ -398,15 +398,15 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
 	struct mod_hdcp_display *display = &hdcp_work[link_index].display;
 	struct mod_hdcp_link *link = &hdcp_work[link_index].link;
 
-	memset(display, 0, sizeof(*display));
-	memset(link, 0, sizeof(*link));
-
-	display->index = aconnector->base.index;
-
 	if (config->dpms_off) {
 		hdcp_remove_display(hdcp_work, link_index, aconnector);
 		return;
 	}
+
+	memset(display, 0, sizeof(*display));
+	memset(link, 0, sizeof(*link));
+
+	display->index = aconnector->base.index;
 	display->state = MOD_HDCP_DISPLAY_ACTIVE;
 
 	if (aconnector->dc_sink != NULL)