16 files changed, 103 insertions, 47 deletions

diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c
index 700050556242..1648887d3f55 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.c
+++ b/drivers/gpu/drm/i915/gvt/execlist.c
@@ -46,6 +46,8 @@
 #define same_context(a, b) (((a)->context_id == (b)->context_id) && \
 		((a)->lrca == (b)->lrca))
 
+static void clean_workloads(struct intel_vgpu *vgpu, unsigned long engine_mask);
+
 static int context_switch_events[] = {
 	[RCS] = RCS_AS_CONTEXT_SWITCH,
 	[BCS] = BCS_AS_CONTEXT_SWITCH,
@@ -499,10 +501,10 @@ static void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 static int complete_execlist_workload(struct intel_vgpu_workload *workload)
 {
 	struct intel_vgpu *vgpu = workload->vgpu;
-	struct intel_vgpu_execlist *execlist =
-		&vgpu->execlist[workload->ring_id];
+	int ring_id = workload->ring_id;
+	struct intel_vgpu_execlist *execlist = &vgpu->execlist[ring_id];
 	struct intel_vgpu_workload *next_workload;
-	struct list_head *next = workload_q_head(vgpu, workload->ring_id)->next;
+	struct list_head *next = workload_q_head(vgpu, ring_id)->next;
 	bool lite_restore = false;
 	int ret;
 
@@ -512,10 +514,25 @@ static int complete_execlist_workload(struct intel_vgpu_workload *workload)
 	release_shadow_batch_buffer(workload);
 	release_shadow_wa_ctx(&workload->wa_ctx);
 
-	if (workload->status || vgpu->resetting)
+	if (workload->status || (vgpu->resetting_eng & ENGINE_MASK(ring_id))) {
+		/* if workload->status is not successful means HW GPU
+		 * has occurred GPU hang or something wrong with i915/GVT,
+		 * and GVT won't inject context switch interrupt to guest.
+		 * So this error is a vGPU hang actually to the guest.
+		 * According to this we should emunlate a vGPU hang. If
+		 * there are pending workloads which are already submitted
+		 * from guest, we should clean them up like HW GPU does.
+		 *
+		 * if it is in middle of engine resetting, the pending
+		 * workloads won't be submitted to HW GPU and will be
+		 * cleaned up during the resetting process later, so doing
+		 * the workload clean up here doesn't have any impact.
+		 **/
+		clean_workloads(vgpu, ENGINE_MASK(ring_id));
 		goto out;
+	}
 
-	if (!list_empty(workload_q_head(vgpu, workload->ring_id))) {
+	if (!list_empty(workload_q_head(vgpu, ring_id))) {
 		struct execlist_ctx_descriptor_format *this_desc, *next_desc;
 
 		next_workload = container_of(next,
diff --git a/drivers/gpu/drm/i915/gvt/firmware.c b/drivers/gpu/drm/i915/gvt/firmware.c
index 5dad9298b2d5..a26c1705430e 100644
--- a/drivers/gpu/drm/i915/gvt/firmware.c
+++ b/drivers/gpu/drm/i915/gvt/firmware.c
@@ -72,11 +72,13 @@ static int expose_firmware_sysfs(struct intel_gvt *gvt)
 	struct intel_gvt_device_info *info = &gvt->device_info;
 	struct pci_dev *pdev = gvt->dev_priv->drm.pdev;
 	struct intel_gvt_mmio_info *e;
+	struct gvt_mmio_block *block = gvt->mmio.mmio_block;
+	int num = gvt->mmio.num_mmio_block;
 	struct gvt_firmware_header *h;
 	void *firmware;
 	void *p;
 	unsigned long size, crc32_start;
-	int i;
+	int i, j;
 	int ret;
 
 	size = sizeof(*h) + info->mmio_size + info->cfg_space_size;
@@ -105,6 +107,13 @@ static int expose_firmware_sysfs(struct intel_gvt *gvt)
 	hash_for_each(gvt->mmio.mmio_info_table, i, e, node)
 		*(u32 *)(p + e->offset) = I915_READ_NOTRACE(_MMIO(e->offset));
 
+	for (i = 0; i < num; i++, block++) {
+		for (j = 0; j < block->size; j += 4)
+			*(u32 *)(p + INTEL_GVT_MMIO_OFFSET(block->offset) + j) =
+				I915_READ_NOTRACE(_MMIO(INTEL_GVT_MMIO_OFFSET(
+							block->offset) + j));
+	}
+
 	memcpy(gvt->firmware.mmio, p, info->mmio_size);
 
 	crc32_start = offsetof(struct gvt_firmware_header, crc32) + 4;
diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h
index 3a74e79eac2f..2964a4d01a66 100644
--- a/drivers/gpu/drm/i915/gvt/gvt.h
+++ b/drivers/gpu/drm/i915/gvt/gvt.h
@@ -149,7 +149,7 @@ struct intel_vgpu {
 	bool active;
 	bool pv_notified;
 	bool failsafe;
-	bool resetting;
+	unsigned int resetting_eng;
 	void *sched_data;
 	struct vgpu_sched_ctl sched_ctl;
 
@@ -195,6 +195,15 @@ struct intel_gvt_fence {
 	unsigned long vgpu_allocated_fence_num;
 };
 
+/* Special MMIO blocks. */
+struct gvt_mmio_block {
+	unsigned int device;
+	i915_reg_t   offset;
+	unsigned int size;
+	gvt_mmio_func read;
+	gvt_mmio_func write;
+};
+
 #define INTEL_GVT_MMIO_HASH_BITS 11
 
 struct intel_gvt_mmio {
@@ -214,6 +223,9 @@ struct intel_gvt_mmio {
 /* This reg could be accessed by unaligned address */
 #define F_UNALIGN	(1 << 6)
 
+	struct gvt_mmio_block *mmio_block;
+	unsigned int num_mmio_block;
+
 	DECLARE_HASHTABLE(mmio_info_table, INTEL_GVT_MMIO_HASH_BITS);
 	unsigned int num_tracked_mmio;
 };
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index 17febe830ff6..feed9921b3b3 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -2857,31 +2857,15 @@ static int init_skl_mmio_info(struct intel_gvt *gvt)
 	return 0;
 }
 
-/* Special MMIO blocks. */
-static struct gvt_mmio_block {
-	unsigned int device;
-	i915_reg_t   offset;
-	unsigned int size;
-	gvt_mmio_func read;
-	gvt_mmio_func write;
-} gvt_mmio_blocks[] = {
-	{D_SKL_PLUS, _MMIO(CSR_MMIO_START_RANGE), 0x3000, NULL, NULL},
-	{D_ALL, _MMIO(MCHBAR_MIRROR_BASE_SNB), 0x40000, NULL, NULL},
-	{D_ALL, _MMIO(VGT_PVINFO_PAGE), VGT_PVINFO_SIZE,
-		pvinfo_mmio_read, pvinfo_mmio_write},
-	{D_ALL, LGC_PALETTE(PIPE_A, 0), 1024, NULL, NULL},
-	{D_ALL, LGC_PALETTE(PIPE_B, 0), 1024, NULL, NULL},
-	{D_ALL, LGC_PALETTE(PIPE_C, 0), 1024, NULL, NULL},
-};
-
 static struct gvt_mmio_block *find_mmio_block(struct intel_gvt *gvt,
 					      unsigned int offset)
 {
 	unsigned long device = intel_gvt_get_device_type(gvt);
-	struct gvt_mmio_block *block = gvt_mmio_blocks;
+	struct gvt_mmio_block *block = gvt->mmio.mmio_block;
+	int num = gvt->mmio.num_mmio_block;
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(gvt_mmio_blocks); i++, block++) {
+	for (i = 0; i < num; i++, block++) {
 		if (!(device & block->device))
 			continue;
 		if (offset >= INTEL_GVT_MMIO_OFFSET(block->offset) &&
@@ -2912,6 +2896,17 @@ void intel_gvt_clean_mmio_info(struct intel_gvt *gvt)
 	gvt->mmio.mmio_attribute = NULL;
 }
 
+/* Special MMIO blocks. */
+static struct gvt_mmio_block mmio_blocks[] = {
+	{D_SKL_PLUS, _MMIO(CSR_MMIO_START_RANGE), 0x3000, NULL, NULL},
+	{D_ALL, _MMIO(MCHBAR_MIRROR_BASE_SNB), 0x40000, NULL, NULL},
+	{D_ALL, _MMIO(VGT_PVINFO_PAGE), VGT_PVINFO_SIZE,
+		pvinfo_mmio_read, pvinfo_mmio_write},
+	{D_ALL, LGC_PALETTE(PIPE_A, 0), 1024, NULL, NULL},
+	{D_ALL, LGC_PALETTE(PIPE_B, 0), 1024, NULL, NULL},
+	{D_ALL, LGC_PALETTE(PIPE_C, 0), 1024, NULL, NULL},
+};
+
 /**
  * intel_gvt_setup_mmio_info - setup MMIO information table for GVT device
  * @gvt: GVT device
@@ -2951,6 +2946,9 @@ int intel_gvt_setup_mmio_info(struct intel_gvt *gvt)
 			goto err;
 	}
 
+	gvt->mmio.mmio_block = mmio_blocks;
+	gvt->mmio.num_mmio_block = ARRAY_SIZE(mmio_blocks);
+
 	gvt_dbg_mmio("traced %u virtual mmio registers\n",
 		     gvt->mmio.num_tracked_mmio);
 	return 0;
@@ -3030,7 +3028,7 @@ int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset,
 	gvt_mmio_func func;
 	int ret;
 
-	if (WARN_ON(bytes > 4))
+	if (WARN_ON(bytes > 8))
 		return -EINVAL;
 
 	/*
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 4f7057d62d88..22e08eb2d0b7 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -432,7 +432,8 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 
 		i915_gem_request_put(fetch_and_zero(&workload->req));
 
-		if (!workload->status && !vgpu->resetting) {
+		if (!workload->status && !(vgpu->resetting_eng &
+					   ENGINE_MASK(ring_id))) {
 			update_guest_context(workload);
 
 			for_each_set_bit(event, workload->pending_events,
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
index 90c14e6e3ea0..3deadcbd5a24 100644
--- a/drivers/gpu/drm/i915/gvt/vgpu.c
+++ b/drivers/gpu/drm/i915/gvt/vgpu.c
@@ -480,11 +480,13 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
 {
 	struct intel_gvt *gvt = vgpu->gvt;
 	struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
+	unsigned int resetting_eng = dmlr ? ALL_ENGINES : engine_mask;
 
 	gvt_dbg_core("------------------------------------------\n");
 	gvt_dbg_core("resseting vgpu%d, dmlr %d, engine_mask %08x\n",
 		     vgpu->id, dmlr, engine_mask);
-	vgpu->resetting = true;
+
+	vgpu->resetting_eng = resetting_eng;
 
 	intel_vgpu_stop_schedule(vgpu);
 	/*
@@ -497,7 +499,7 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
 		mutex_lock(&gvt->lock);
 	}
 
-	intel_vgpu_reset_execlist(vgpu, dmlr ? ALL_ENGINES : engine_mask);
+	intel_vgpu_reset_execlist(vgpu, resetting_eng);
 
 	/* full GPU reset or device model level reset */
 	if (engine_mask == ALL_ENGINES || dmlr) {
@@ -520,7 +522,7 @@ void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
 		}
 	}
 
-	vgpu->resetting = false;
+	vgpu->resetting_eng = 0;
 	gvt_dbg_core("reset vgpu%d done\n", vgpu->id);
 	gvt_dbg_core("------------------------------------------\n");
 }
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 00d8967c8512..d1bd53b73738 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -4580,7 +4580,7 @@ static void gen9_sseu_device_status(struct drm_i915_private *dev_priv,
 
 		sseu->slice_mask |= BIT(s);
 
-		if (IS_GEN9_BC(dev_priv))
+		if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv))
 			sseu->subslice_mask =
 				INTEL_INFO(dev_priv)->sseu.subslice_mask;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 39ed58a21fc1..e1e971ee2ed5 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -688,19 +688,19 @@ static inline bool skip_rcs_switch(struct i915_hw_ppgtt *ppgtt,
 }
 
 static bool
-needs_pd_load_pre(struct i915_hw_ppgtt *ppgtt,
-		  struct intel_engine_cs *engine,
-		  struct i915_gem_context *to)
+needs_pd_load_pre(struct i915_hw_ppgtt *ppgtt, struct intel_engine_cs *engine)
 {
+	struct i915_gem_context *from = engine->legacy_active_context;
+
 	if (!ppgtt)
 		return false;
 
 	/* Always load the ppgtt on first use */
-	if (!engine->legacy_active_context)
+	if (!from)
 		return true;
 
 	/* Same context without new entries, skip */
-	if (engine->legacy_active_context == to &&
+	if ((!from->ppgtt || from->ppgtt == ppgtt) &&
 	    !(intel_engine_flag(engine) & ppgtt->pd_dirty_rings))
 		return false;
 
@@ -744,7 +744,7 @@ static int do_rcs_switch(struct drm_i915_gem_request *req)
 	if (skip_rcs_switch(ppgtt, engine, to))
 		return 0;
 
-	if (needs_pd_load_pre(ppgtt, engine, to)) {
+	if (needs_pd_load_pre(ppgtt, engine)) {
 		/* Older GENs and non render rings still want the load first,
 		 * "PP_DCLV followed by PP_DIR_BASE register through Load
 		 * Register Immediate commands in Ring Buffer before submitting
@@ -841,7 +841,7 @@ int i915_switch_context(struct drm_i915_gem_request *req)
 		struct i915_hw_ppgtt *ppgtt =
 			to->ppgtt ?: req->i915->mm.aliasing_ppgtt;
 
-		if (needs_pd_load_pre(ppgtt, engine, to)) {
+		if (needs_pd_load_pre(ppgtt, engine)) {
 			int ret;
 
 			trace_switch_mm(engine, to);
@@ -852,6 +852,7 @@ int i915_switch_context(struct drm_i915_gem_request *req)
 			ppgtt->pd_dirty_rings &= ~intel_engine_flag(engine);
 		}
 
+		engine->legacy_active_context = to;
 		return 0;
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index 7032c542a9b1..4dd4c2159a92 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -242,6 +242,10 @@ int i915_gem_render_state_emit(struct drm_i915_gem_request *req)
 			goto err_unpin;
 	}
 
+	ret = req->engine->emit_flush(req, EMIT_INVALIDATE);
+	if (ret)
+		goto err_unpin;
+
 	ret = req->engine->emit_bb_start(req,
 					 so->batch_offset, so->batch_size,
 					 I915_DISPATCH_SECURE);
diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
index 1032f98add11..77fb39808131 100644
--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
@@ -43,16 +43,21 @@ static bool shrinker_lock(struct drm_i915_private *dev_priv, bool *unlock)
 		return true;
 
 	case MUTEX_TRYLOCK_FAILED:
+		*unlock = false;
+		preempt_disable();
 		do {
 			cpu_relax();
 			if (mutex_trylock(&dev_priv->drm.struct_mutex)) {
-	case MUTEX_TRYLOCK_SUCCESS:
 				*unlock = true;
-				return true;
+				break;
 			}
 		} while (!need_resched());
+		preempt_enable();
+		return *unlock;
 
-		return false;
+	case MUTEX_TRYLOCK_SUCCESS:
+		*unlock = true;
+		return true;
 	}
 
 	BUG();
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 9cd22f83b0cf..f33d90226704 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1601,11 +1601,11 @@ static int gen8_emit_oa_config(struct drm_i915_gem_request *req)
 	u32 *cs;
 	int i;
 
-	cs = intel_ring_begin(req, n_flex_regs * 2 + 4);
+	cs = intel_ring_begin(req, ARRAY_SIZE(flex_mmio) * 2 + 4);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
-	*cs++ = MI_LOAD_REGISTER_IMM(n_flex_regs + 1);
+	*cs++ = MI_LOAD_REGISTER_IMM(ARRAY_SIZE(flex_mmio) + 1);
 
 	*cs++ = i915_mmio_reg_offset(GEN8_OACTXCONTROL);
 	*cs++ = (dev_priv->perf.oa.period_exponent << GEN8_OA_TIMER_PERIOD_SHIFT) |
diff --git a/drivers/gpu/drm/i915/intel_color.c b/drivers/gpu/drm/i915/intel_color.c
index 306c6b06b330..17c4ae7e4e7c 100644
--- a/drivers/gpu/drm/i915/intel_color.c
+++ b/drivers/gpu/drm/i915/intel_color.c
@@ -398,6 +398,7 @@ static void bdw_load_gamma_lut(struct drm_crtc_state *state, u32 offset)
 		}
 
 		/* Program the max register to clamp values > 1.0. */
+		i = lut_size - 1;
 		I915_WRITE(PREC_PAL_GC_MAX(pipe, 0),
 			   drm_color_lut_extract(lut[i].red, 16));
 		I915_WRITE(PREC_PAL_GC_MAX(pipe, 1),
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 9edeaaef77ad..d3b3252a8742 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -1762,7 +1762,7 @@ cnl_get_buf_trans_edp(struct drm_i915_private *dev_priv,
 	if (dev_priv->vbt.edp.low_vswing) {
 		if (voltage == VOLTAGE_INFO_0_85V) {
 			*n_entries = ARRAY_SIZE(cnl_ddi_translations_edp_0_85V);
-			return cnl_ddi_translations_dp_0_85V;
+			return cnl_ddi_translations_edp_0_85V;
 		} else if (voltage == VOLTAGE_INFO_0_95V) {
 			*n_entries = ARRAY_SIZE(cnl_ddi_translations_edp_0_95V);
 			return cnl_ddi_translations_edp_0_95V;
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 9471c88d449e..cc484b56eeaa 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3485,6 +3485,13 @@ void intel_prepare_reset(struct drm_i915_private *dev_priv)
 	    !gpu_reset_clobbers_display(dev_priv))
 		return;
 
+	/* We have a modeset vs reset deadlock, defensively unbreak it.
+	 *
+	 * FIXME: We can do a _lot_ better, this is just a first iteration.
+	 */
+	i915_gem_set_wedged(dev_priv);
+	DRM_DEBUG_DRIVER("Wedging GPU to avoid deadlocks with pending modeset updates\n");
+
 	/*
 	 * Need mode_config.mutex so that we don't
 	 * trample ongoing ->detect() and whatnot.
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index 52b3a1fd4059..57ef5833c427 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -63,7 +63,6 @@ enum {
 };
 
 /* Logical Rings */
-void intel_logical_ring_stop(struct intel_engine_cs *engine);
 void intel_logical_ring_cleanup(struct intel_engine_cs *engine);
 int logical_render_ring_init(struct intel_engine_cs *engine);
 int logical_xcs_ring_init(struct intel_engine_cs *engine);
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 96c2cbd81869..593349be8b9d 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -469,7 +469,7 @@ static u32 intel_panel_compute_brightness(struct intel_connector *connector,
 
 	if (i915.invert_brightness > 0 ||
 	    dev_priv->quirks & QUIRK_INVERT_BRIGHTNESS) {
-		return panel->backlight.max - val;
+		return panel->backlight.max - val + panel->backlight.min;
 	}
 
 	return val;