Merge tag 'drm-intel-next-2018-05-14' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

Last drm/i915 changes for v4.18:

- NV12 enabling (Chandra, Maarten)
- ICL workarounds (Oscar)
- ICL basic DPLL enabling (Paulo)
- GVT updates
- DP link config refactoring (Jani)
- Module parameter to override DMC firmware (Jani)
- PSR updates (José, DK, Daniel, Ville)
- ICL DP vswing programming (Manasi)
- ICL DBuf slice updates (Mahesh)
- Selftest fixes and updates (Chris, Matthew, Oscar)
- Execlist fixes and updates (Chris)
- Stolen memory first 4k fix (Hans de Goede)
- wait_for fixes (Mika)
- Tons of GEM improvements (Chris)
- Plenty of other fixes and improvements (Everyone)
- Crappy changelog (Me)
Signed-off-by: NDave Airlie <airlied@redhat.com>

# gpg: Signature made Mon 14 May 2018 11:04:24 PM AEST
# gpg:                using RSA key D398079D26ABEE6F
# gpg: Good signature from "Jani Nikula <jani.nikula@intel.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 1565 A65B 77B0 632E 1124  E59C D398 079D 26AB EE6F

# Conflicts:
#	drivers/gpu/drm/i915/intel_lrc.c
#	drivers/gpu/drm/i915/intel_sprite.c
Link: https://patchwork.freedesktop.org/patch/msgid/87k1s51bvw.fsf@intel.com

drivers/gpu/drm/drm_atomic.c

@@ -1716,11 +1716,15 @@ int drm_atomic_check_only(struct drm_atomic_state *state)
 		}
 	}
 
-	if (config->funcs->atomic_check)
+	if (config->funcs->atomic_check) {
 		ret = config->funcs->atomic_check(state->dev, state);
 
-	if (ret)
-		return ret;
+		if (ret) {
+			DRM_DEBUG_ATOMIC("atomic driver check for %p failed: %d\n",
+					 state, ret);
+			return ret;
+		}
+	}
 
 	if (!state->allow_modeset) {
 		for_each_new_crtc_in_state(state, crtc, crtc_state, i) {

drivers/gpu/drm/i915/Makefile

@@ -18,6 +18,10 @@ subdir-ccflags-y += $(call cc-disable-warning, type-limits)
 subdir-ccflags-y += $(call cc-disable-warning, missing-field-initializers)
 subdir-ccflags-y += $(call cc-disable-warning, implicit-fallthrough)
 subdir-ccflags-y += $(call cc-disable-warning, unused-but-set-variable)
+# clang warnings
+subdir-ccflags-y += $(call cc-disable-warning, sign-compare)
+subdir-ccflags-y += $(call cc-disable-warning, sometimes-uninitialized)
+subdir-ccflags-y += $(call cc-disable-warning, initializer-overrides)
 subdir-ccflags-$(CONFIG_DRM_I915_WERROR) += -Werror
 
 # Fine grained warnings disable
@@ -67,11 +71,11 @@ i915-y += i915_cmd_parser.o \
 	  i915_gem_shrinker.o \
 	  i915_gem_stolen.o \
 	  i915_gem_tiling.o \
-	  i915_gem_timeline.o \
 	  i915_gem_userptr.o \
 	  i915_gemfs.o \
 	  i915_query.o \
 	  i915_request.o \
+	  i915_timeline.o \
 	  i915_trace_points.o \
 	  i915_vma.o \
 	  intel_breadcrumbs.o \
@@ -154,7 +158,8 @@ i915-y += dvo_ch7017.o \
 i915-$(CONFIG_DRM_I915_CAPTURE_ERROR) += i915_gpu_error.o
 i915-$(CONFIG_DRM_I915_SELFTEST) += \
 	selftests/i915_random.o \
-	selftests/i915_selftest.o
+	selftests/i915_selftest.o \
+	selftests/igt_flush_test.o
 
 # virtual gpu code
 i915-y += i915_vgpu.o

drivers/gpu/drm/i915/gvt/cmd_parser.c

@@ -813,15 +813,31 @@ static inline bool is_force_nonpriv_mmio(unsigned int offset)
 }
 
 static int force_nonpriv_reg_handler(struct parser_exec_state *s,
-				     unsigned int offset, unsigned int index)
+		unsigned int offset, unsigned int index, char *cmd)
 {
 	struct intel_gvt *gvt = s->vgpu->gvt;
-	unsigned int data = cmd_val(s, index + 1);
+	unsigned int data;
+	u32 ring_base;
+	u32 nopid;
+	struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv;
+
+	if (!strcmp(cmd, "lri"))
+		data = cmd_val(s, index + 1);
+	else {
+		gvt_err("Unexpected forcenonpriv 0x%x write from cmd %s\n",
+			offset, cmd);
+		return -EINVAL;
+	}
+
+	ring_base = dev_priv->engine[s->ring_id]->mmio_base;
+	nopid = i915_mmio_reg_offset(RING_NOPID(ring_base));
 
-	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data)) {
+	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data) &&
+			data != nopid) {
 		gvt_err("Unexpected forcenonpriv 0x%x LRI write, value=0x%x\n",
 			offset, data);
-		return -EPERM;
+		patch_value(s, cmd_ptr(s, index), nopid);
+		return 0;
 	}
 	return 0;
 }
@@ -869,7 +885,7 @@ static int cmd_reg_handler(struct parser_exec_state *s,
 		return -EINVAL;
 
 	if (is_force_nonpriv_mmio(offset) &&
-		force_nonpriv_reg_handler(s, offset, index))
+		force_nonpriv_reg_handler(s, offset, index, cmd))
 		return -EPERM;
 
 	if (offset == i915_mmio_reg_offset(DERRMR) ||
@@ -1604,7 +1620,8 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s)
 	if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)
 		|| IS_KABYLAKE(gvt->dev_priv)) {
 		/* BDW decides privilege based on address space */
-		if (cmd_val(s, 0) & (1 << 8))
+		if (cmd_val(s, 0) & (1 << 8) &&
+			!(s->vgpu->scan_nonprivbb & (1 << s->ring_id)))
 			return 0;
 	}
 	return 1;
@@ -1618,6 +1635,8 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size)
 	bool bb_end = false;
 	struct intel_vgpu *vgpu = s->vgpu;
 	u32 cmd;
+	struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ?
+		s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm;
 
 	*bb_size = 0;
 
@@ -1629,18 +1648,22 @@ static int find_bb_size(struct parser_exec_state *s, unsigned long *bb_size)
 	cmd = cmd_val(s, 0);
 	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 		return -EBADRQC;
 	}
 	do {
-		if (copy_gma_to_hva(s->vgpu, s->vgpu->gtt.ggtt_mm,
+		if (copy_gma_to_hva(s->vgpu, mm,
 				gma, gma + 4, &cmd) < 0)
 			return -EFAULT;
 		info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 		if (info == NULL) {
-			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+			gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 			return -EBADRQC;
 		}
 
@@ -1666,6 +1689,9 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	unsigned long gma = 0;
 	unsigned long bb_size;
 	int ret = 0;
+	struct intel_vgpu_mm *mm = (s->buf_addr_type == GTT_BUFFER) ?
+		s->vgpu->gtt.ggtt_mm : s->workload->shadow_mm;
+	unsigned long gma_start_offset = 0;
 
 	/* get the start gm address of the batch buffer */
 	gma = get_gma_bb_from_cmd(s, 1);
@@ -1680,8 +1706,24 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	if (!bb)
 		return -ENOMEM;
 
+	bb->ppgtt = (s->buf_addr_type == GTT_BUFFER) ? false : true;
+
+	/* the gma_start_offset stores the batch buffer's start gma's
+	 * offset relative to page boundary. so for non-privileged batch
+	 * buffer, the shadowed gem object holds exactly the same page
+	 * layout as original gem object. This is for the convience of
+	 * replacing the whole non-privilged batch buffer page to this
+	 * shadowed one in PPGTT at the same gma address. (this replacing
+	 * action is not implemented yet now, but may be necessary in
+	 * future).
+	 * for prileged batch buffer, we just change start gma address to
+	 * that of shadowed page.
+	 */
+	if (bb->ppgtt)
+		gma_start_offset = gma & ~I915_GTT_PAGE_MASK;
+
 	bb->obj = i915_gem_object_create(s->vgpu->gvt->dev_priv,
-					 roundup(bb_size, PAGE_SIZE));
+			 roundup(bb_size + gma_start_offset, PAGE_SIZE));
 	if (IS_ERR(bb->obj)) {
 		ret = PTR_ERR(bb->obj);
 		goto err_free_bb;
@@ -1702,9 +1744,9 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 		bb->clflush &= ~CLFLUSH_BEFORE;
 	}
 
-	ret = copy_gma_to_hva(s->vgpu, s->vgpu->gtt.ggtt_mm,
+	ret = copy_gma_to_hva(s->vgpu, mm,
 			      gma, gma + bb_size,
-			      bb->va);
+			      bb->va + gma_start_offset);
 	if (ret < 0) {
 		gvt_vgpu_err("fail to copy guest ring buffer\n");
 		ret = -EFAULT;
@@ -1730,7 +1772,7 @@ static int perform_bb_shadow(struct parser_exec_state *s)
 	 * buffer's gma in pair. After all, we don't want to pin the shadow
 	 * buffer here (too early).
 	 */
-	s->ip_va = bb->va;
+	s->ip_va = bb->va + gma_start_offset;
 	s->ip_gma = gma;
 	return 0;
 err_unmap:
@@ -2469,15 +2511,18 @@ static int cmd_parser_exec(struct parser_exec_state *s)
 
 	info = get_cmd_info(s->vgpu->gvt, cmd, s->ring_id);
 	if (info == NULL) {
-		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x\n",
-				cmd, get_opcode(cmd, s->ring_id));
+		gvt_vgpu_err("unknown cmd 0x%x, opcode=0x%x, addr_type=%s, ring %d, workload=%p\n",
+				cmd, get_opcode(cmd, s->ring_id),
+				(s->buf_addr_type == PPGTT_BUFFER) ?
+				"ppgtt" : "ggtt", s->ring_id, s->workload);
 		return -EBADRQC;
 	}
 
 	s->info = info;
 
 	trace_gvt_command(vgpu->id, s->ring_id, s->ip_gma, s->ip_va,
-			  cmd_length(s), s->buf_type);
+			  cmd_length(s), s->buf_type, s->buf_addr_type,
+			  s->workload, info->name);
 
 	if (info->handler) {
 		ret = info->handler(s);

drivers/gpu/drm/i915/gvt/debugfs.c

@@ -124,6 +124,68 @@ static int vgpu_mmio_diff_show(struct seq_file *s, void *unused)
 }
 DEFINE_SHOW_ATTRIBUTE(vgpu_mmio_diff);
 
+static int
+vgpu_scan_nonprivbb_get(void *data, u64 *val)
+{
+	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
+	*val = vgpu->scan_nonprivbb;
+	return 0;
+}
+
+/*
+ * set/unset bit engine_id of vgpu->scan_nonprivbb to turn on/off scanning
+ * of non-privileged batch buffer. e.g.
+ * if vgpu->scan_nonprivbb=3, then it will scan non-privileged batch buffer
+ * on engine 0 and 1.
+ */
+static int
+vgpu_scan_nonprivbb_set(void *data, u64 val)
+{
+	struct intel_vgpu *vgpu = (struct intel_vgpu *)data;
+	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
+	enum intel_engine_id id;
+	char buf[128], *s;
+	int len;
+
+	val &= (1 << I915_NUM_ENGINES) - 1;
+
+	if (vgpu->scan_nonprivbb == val)
+		return 0;
+
+	if (!val)
+		goto done;
+
+	len = sprintf(buf,
+		"gvt: vgpu %d turns on non-privileged batch buffers scanning on Engines:",
+		vgpu->id);
+
+	s = buf + len;
+
+	for (id = 0; id < I915_NUM_ENGINES; id++) {
+		struct intel_engine_cs *engine;
+
+		engine = dev_priv->engine[id];
+		if (engine && (val & (1 << id))) {
+			len = snprintf(s, 4, "%d, ", engine->id);
+			s += len;
+		} else
+			val &=  ~(1 << id);
+	}
+
+	if (val)
+		sprintf(s, "low performance expected.");
+
+	pr_warn("%s\n", buf);
+
+done:
+	vgpu->scan_nonprivbb = val;
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(vgpu_scan_nonprivbb_fops,
+			vgpu_scan_nonprivbb_get, vgpu_scan_nonprivbb_set,
+			"0x%llx\n");
+
 /**
  * intel_gvt_debugfs_add_vgpu - register debugfs entries for a vGPU
  * @vgpu: a vGPU
@@ -151,6 +213,11 @@ int intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu)
 	if (!ent)
 		return -ENOMEM;
 
+	ent = debugfs_create_file("scan_nonprivbb", 0644, vgpu->debugfs,
+				 vgpu, &vgpu_scan_nonprivbb_fops);
+	if (!ent)
+		return -ENOMEM;
+
 	return 0;
 }
 

drivers/gpu/drm/i915/gvt/gvt.h

@@ -99,7 +99,6 @@ struct intel_vgpu_fence {
 struct intel_vgpu_mmio {
 	void *vreg;
 	void *sreg;
-	bool disable_warn_untrack;
 };
 
 #define INTEL_GVT_MAX_BAR_NUM 4
@@ -226,6 +225,7 @@ struct intel_vgpu {
 
 	struct completion vblank_done;
 
+	u32 scan_nonprivbb;
 };
 
 /* validating GM healthy status*/

drivers/gpu/drm/i915/gvt/handlers.c

@@ -191,6 +191,8 @@ static int sanitize_fence_mmio_access(struct intel_vgpu *vgpu,
 	unsigned int max_fence = vgpu_fence_sz(vgpu);
 
 	if (fence_num >= max_fence) {
+		gvt_vgpu_err("access oob fence reg %d/%d\n",
+			     fence_num, max_fence);
 
 		/* When guest access oob fence regs without access
 		 * pv_info first, we treat guest not supporting GVT,
@@ -200,11 +202,6 @@ static int sanitize_fence_mmio_access(struct intel_vgpu *vgpu,
 			enter_failsafe_mode(vgpu,
 					GVT_FAILSAFE_UNSUPPORTED_GUEST);
 
-		if (!vgpu->mmio.disable_warn_untrack) {
-			gvt_vgpu_err("found oob fence register access\n");
-			gvt_vgpu_err("total fence %d, access fence %d\n",
-				     max_fence, fence_num);
-		}
 		memset(p_data, 0, bytes);
 		return -EINVAL;
 	}
@@ -477,22 +474,28 @@ static int force_nonpriv_write(struct intel_vgpu *vgpu,
 	unsigned int offset, void *p_data, unsigned int bytes)
 {
 	u32 reg_nonpriv = *(u32 *)p_data;
+	int ring_id = intel_gvt_render_mmio_to_ring_id(vgpu->gvt, offset);
+	u32 ring_base;
+	struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv;
 	int ret = -EINVAL;
 
-	if ((bytes != 4) || ((offset & (bytes - 1)) != 0)) {
-		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV offset %x(%dB)\n",
-			vgpu->id, offset, bytes);
+	if ((bytes != 4) || ((offset & (bytes - 1)) != 0) || ring_id < 0) {
+		gvt_err("vgpu(%d) ring %d Invalid FORCE_NONPRIV offset %x(%dB)\n",
+			vgpu->id, ring_id, offset, bytes);
 		return ret;
 	}
 
-	if (in_whitelist(reg_nonpriv)) {
+	ring_base = dev_priv->engine[ring_id]->mmio_base;
+
+	if (in_whitelist(reg_nonpriv) ||
+		reg_nonpriv == i915_mmio_reg_offset(RING_NOPID(ring_base))) {
 		ret = intel_vgpu_default_mmio_write(vgpu, offset, p_data,
 			bytes);
-	} else {
-		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x\n",
-			vgpu->id, reg_nonpriv);
-	}
-	return ret;
+	} else
+		gvt_err("vgpu(%d) Invalid FORCE_NONPRIV write %x at offset %x\n",
+			vgpu->id, reg_nonpriv, offset);
+
+	return 0;
 }
 
 static int ddi_buf_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
@@ -3092,9 +3095,7 @@ int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset,
 	 */
 	mmio_info = find_mmio_info(gvt, offset);
 	if (!mmio_info) {
-		if (!vgpu->mmio.disable_warn_untrack)
-			gvt_vgpu_err("untracked MMIO %08x len %d\n",
-				     offset, bytes);
+		gvt_dbg_mmio("untracked MMIO %08x len %d\n", offset, bytes);
 		goto default_rw;
 	}
 

drivers/gpu/drm/i915/gvt/mmio.c

@@ -244,8 +244,6 @@ void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu, bool dmlr)
 
 		/* set the bit 0:2(Core C-State ) to C0 */
 		vgpu_vreg_t(vgpu, GEN6_GT_CORE_STATUS) = 0;
-
-		vgpu->mmio.disable_warn_untrack = false;
 	} else {
 #define GVT_GEN8_MMIO_RESET_OFFSET		(0x44200)
 		/* only reset the engine related, so starting with 0x44200

drivers/gpu/drm/i915/gvt/mmio_context.c

@@ -448,7 +448,7 @@ static void switch_mocs(struct intel_vgpu *pre, struct intel_vgpu *next,
 
 bool is_inhibit_context(struct i915_gem_context *ctx, int ring_id)
 {
-	u32 *reg_state = ctx->engine[ring_id].lrc_reg_state;
+	u32 *reg_state = ctx->__engine[ring_id].lrc_reg_state;
 	u32 inhibit_mask =
 		_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT);
 

drivers/gpu/drm/i915/gvt/sched_policy.c

@@ -53,7 +53,6 @@ struct vgpu_sched_data {
 	bool active;
 
 	ktime_t sched_in_time;
-	ktime_t sched_out_time;
 	ktime_t sched_time;
 	ktime_t left_ts;
 	ktime_t allocated_ts;
@@ -66,17 +65,22 @@ struct gvt_sched_data {
 	struct hrtimer timer;
 	unsigned long period;
 	struct list_head lru_runq_head;
+	ktime_t expire_time;
 };
 
-static void vgpu_update_timeslice(struct intel_vgpu *pre_vgpu)
+static void vgpu_update_timeslice(struct intel_vgpu *vgpu, ktime_t cur_time)
 {
 	ktime_t delta_ts;
-	struct vgpu_sched_data *vgpu_data = pre_vgpu->sched_data;
+	struct vgpu_sched_data *vgpu_data;
 
-	delta_ts = vgpu_data->sched_out_time - vgpu_data->sched_in_time;
+	if (!vgpu || vgpu == vgpu->gvt->idle_vgpu)
+		return;
 
-	vgpu_data->sched_time += delta_ts;
-	vgpu_data->left_ts -= delta_ts;
+	vgpu_data = vgpu->sched_data;
+	delta_ts = ktime_sub(cur_time, vgpu_data->sched_in_time);
+	vgpu_data->sched_time = ktime_add(vgpu_data->sched_time, delta_ts);
+	vgpu_data->left_ts = ktime_sub(vgpu_data->left_ts, delta_ts);
+	vgpu_data->sched_in_time = cur_time;
 }
 
 #define GVT_TS_BALANCE_PERIOD_MS 100
@@ -150,11 +154,7 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt)
 	}
 
 	cur_time = ktime_get();
-	if (scheduler->current_vgpu) {
-		vgpu_data = scheduler->current_vgpu->sched_data;
-		vgpu_data->sched_out_time = cur_time;
-		vgpu_update_timeslice(scheduler->current_vgpu);
-	}
+	vgpu_update_timeslice(scheduler->current_vgpu, cur_time);
 	vgpu_data = scheduler->next_vgpu->sched_data;
 	vgpu_data->sched_in_time = cur_time;
 
@@ -226,17 +226,22 @@ static void tbs_sched_func(struct gvt_sched_data *sched_data)
 void intel_gvt_schedule(struct intel_gvt *gvt)
 {
 	struct gvt_sched_data *sched_data = gvt->scheduler.sched_data;
-	static uint64_t timer_check;
+	ktime_t cur_time;
 
 	mutex_lock(&gvt->lock);
+	cur_time = ktime_get();
 
 	if (test_and_clear_bit(INTEL_GVT_REQUEST_SCHED,
 				(void *)&gvt->service_request)) {
-		if (!(timer_check++ % GVT_TS_BALANCE_PERIOD_MS))
+		if (cur_time >= sched_data->expire_time) {
 			gvt_balance_timeslice(sched_data);
+			sched_data->expire_time = ktime_add_ms(
+				cur_time, GVT_TS_BALANCE_PERIOD_MS);
+		}
 	}
 	clear_bit(INTEL_GVT_REQUEST_EVENT_SCHED, (void *)&gvt->service_request);
 
+	vgpu_update_timeslice(gvt->scheduler.current_vgpu, cur_time);
 	tbs_sched_func(sched_data);
 
 	mutex_unlock(&gvt->lock);

drivers/gpu/drm/i915/gvt/scheduler.c

@@ -58,7 +58,7 @@ static void update_shadow_pdps(struct intel_vgpu_workload *workload)
 	int ring_id = workload->ring_id;
 	struct i915_gem_context *shadow_ctx = vgpu->submission.shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 
@@ -97,7 +97,7 @@ static void sr_oa_regs(struct intel_vgpu_workload *workload,
 		i915_mmio_reg_offset(EU_PERF_CNTL6),
 	};
 
-	if (!workload || !reg_state || workload->ring_id != RCS)
+	if (workload->ring_id != RCS)
 		return;
 
 	if (save) {
@@ -130,7 +130,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	int ring_id = workload->ring_id;
 	struct i915_gem_context *shadow_ctx = vgpu->submission.shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *dst;
@@ -283,7 +283,7 @@ static int shadow_context_status_change(struct notifier_block *nb,
 static void shadow_context_descriptor_update(struct i915_gem_context *ctx,
 		struct intel_engine_cs *engine)
 {
-	struct intel_context *ce = &ctx->engine[engine->id];
+	struct intel_context *ce = to_intel_context(ctx, engine);
 	u64 desc = 0;
 
 	desc = ce->lrc_desc;
@@ -389,7 +389,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	 * shadow_ctx pages invalid. So gvt need to pin itself. After update
 	 * the guest context, gvt can unpin the shadow_ctx safely.
 	 */
-	ring = engine->context_pin(engine, shadow_ctx);
+	ring = intel_context_pin(shadow_ctx, engine);
 	if (IS_ERR(ring)) {
 		ret = PTR_ERR(ring);
 		gvt_vgpu_err("fail to pin shadow context\n");
@@ -403,7 +403,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	return 0;
 
 err_unpin:
-	engine->context_unpin(engine, shadow_ctx);
+	intel_context_unpin(shadow_ctx, engine);
 err_shadow:
 	release_shadow_wa_ctx(&workload->wa_ctx);
 err_scan:
@@ -437,7 +437,7 @@ static int intel_gvt_generate_request(struct intel_vgpu_workload *workload)
 	return 0;
 
 err_unpin:
-	engine->context_unpin(engine, shadow_ctx);
+	intel_context_unpin(shadow_ctx, engine);
 	release_shadow_wa_ctx(&workload->wa_ctx);
 	return ret;
 }
@@ -452,12 +452,6 @@ static int prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 	int ret;
 
 	list_for_each_entry(bb, &workload->shadow_bb, list) {
-		bb->vma = i915_gem_object_ggtt_pin(bb->obj, NULL, 0, 0, 0);
-		if (IS_ERR(bb->vma)) {
-			ret = PTR_ERR(bb->vma);
-			goto err;
-		}
-
 		/* For privilge batch buffer and not wa_ctx, the bb_start_cmd_va
 		 * is only updated into ring_scan_buffer, not real ring address
 		 * allocated in later copy_workload_to_ring_buffer. pls be noted
@@ -469,25 +463,53 @@ static int prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload)
 			bb->bb_start_cmd_va = workload->shadow_ring_buffer_va
 				+ bb->bb_offset;
 
-		/* relocate shadow batch buffer */
-		bb->bb_start_cmd_va[1] = i915_ggtt_offset(bb->vma);
-		if (gmadr_bytes == 8)
-			bb->bb_start_cmd_va[2] = 0;
+		if (bb->ppgtt) {
+			/* for non-priv bb, scan&shadow is only for
+			 * debugging purpose, so the content of shadow bb
+			 * is the same as original bb. Therefore,
+			 * here, rather than switch to shadow bb's gma
+			 * address, we directly use original batch buffer's
+			 * gma address, and send original bb to hardware
+			 * directly
+			 */
+			if (bb->clflush & CLFLUSH_AFTER) {
+				drm_clflush_virt_range(bb->va,
+						bb->obj->base.size);
+				bb->clflush &= ~CLFLUSH_AFTER;
+			}
+			i915_gem_obj_finish_shmem_access(bb->obj);
+			bb->accessing = false;
+
+		} else {
+			bb->vma = i915_gem_object_ggtt_pin(bb->obj,
+					NULL, 0, 0, 0);
+			if (IS_ERR(bb->vma)) {
+				ret = PTR_ERR(bb->vma);
+				goto err;
+			}
 
-		/* No one is going to touch shadow bb from now on. */
-		if (bb->clflush & CLFLUSH_AFTER) {
-			drm_clflush_virt_range(bb->va, bb->obj->base.size);
-			bb->clflush &= ~CLFLUSH_AFTER;
-		}
+			/* relocate shadow batch buffer */
+			bb->bb_start_cmd_va[1] = i915_ggtt_offset(bb->vma);
+			if (gmadr_bytes == 8)
+				bb->bb_start_cmd_va[2] = 0;
 
-		ret = i915_gem_object_set_to_gtt_domain(bb->obj, false);
-		if (ret)
-			goto err;
+			/* No one is going to touch shadow bb from now on. */
+			if (bb->clflush & CLFLUSH_AFTER) {
+				drm_clflush_virt_range(bb->va,
+						bb->obj->base.size);
+				bb->clflush &= ~CLFLUSH_AFTER;
+			}
 
-		i915_gem_obj_finish_shmem_access(bb->obj);
-		bb->accessing = false;
+			ret = i915_gem_object_set_to_gtt_domain(bb->obj,
+					false);
+			if (ret)
+				goto err;
 
-		i915_vma_move_to_active(bb->vma, workload->req, 0);
+			i915_gem_obj_finish_shmem_access(bb->obj);
+			bb->accessing = false;
+
+			i915_vma_move_to_active(bb->vma, workload->req, 0);
+		}
 	}
 	return 0;
 err:
@@ -504,7 +526,7 @@ static int update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	struct intel_vgpu_submission *s = &workload->vgpu->submission;
 	struct i915_gem_context *shadow_ctx = s->shadow_ctx;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 
@@ -666,7 +688,7 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 
 	ret = prepare_workload(workload);
 	if (ret) {
-		engine->context_unpin(engine, shadow_ctx);
+		intel_context_unpin(shadow_ctx, engine);
 		goto out;
 	}
 
@@ -749,7 +771,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 	struct i915_gem_context *shadow_ctx = s->shadow_ctx;
 	int ring_id = workload->ring_id;
 	struct drm_i915_gem_object *ctx_obj =
-		shadow_ctx->engine[ring_id].state->obj;
+		shadow_ctx->__engine[ring_id].state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *src;
@@ -876,7 +898,7 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 		}
 		mutex_lock(&dev_priv->drm.struct_mutex);
 		/* unpin shadow ctx as the shadow_ctx update is done */
-		engine->context_unpin(engine, s->shadow_ctx);
+		intel_context_unpin(s->shadow_ctx, engine);
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
 
@@ -1134,9 +1156,6 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	if (IS_ERR(s->shadow_ctx))
 		return PTR_ERR(s->shadow_ctx);
 
-	if (HAS_LOGICAL_RING_PREEMPTION(vgpu->gvt->dev_priv))
-		s->shadow_ctx->priority = INT_MAX;
-
 	bitmap_zero(s->shadow_ctx_desc_updated, I915_NUM_ENGINES);
 
 	s->workloads = kmem_cache_create_usercopy("gvt-g_vgpu_workload",

drivers/gpu/drm/i915/gvt/scheduler.h

@@ -125,6 +125,7 @@ struct intel_vgpu_shadow_bb {
 	unsigned int clflush;
 	bool accessing;
 	unsigned long bb_offset;
+	bool ppgtt;
 };
 
 #define workload_q_head(vgpu, ring_id) \

drivers/gpu/drm/i915/gvt/trace.h

@@ -224,19 +224,25 @@ TRACE_EVENT(oos_sync,
 	TP_printk("%s", __entry->buf)
 );
 
+#define GVT_CMD_STR_LEN 40
 TRACE_EVENT(gvt_command,
-	TP_PROTO(u8 vgpu_id, u8 ring_id, u32 ip_gma, u32 *cmd_va, u32 cmd_len,
-		 u32 buf_type),
+	TP_PROTO(u8 vgpu_id, u8 ring_id, u32 ip_gma, u32 *cmd_va,
+		u32 cmd_len,  u32 buf_type, u32 buf_addr_type,
+		void *workload, char *cmd_name),
 
-	TP_ARGS(vgpu_id, ring_id, ip_gma, cmd_va, cmd_len, buf_type),
+	TP_ARGS(vgpu_id, ring_id, ip_gma, cmd_va, cmd_len, buf_type,
+		buf_addr_type, workload, cmd_name),
 
 	TP_STRUCT__entry(
 		__field(u8, vgpu_id)
 		__field(u8, ring_id)
 		__field(u32, ip_gma)
 		__field(u32, buf_type)
+		__field(u32, buf_addr_type)
 		__field(u32, cmd_len)
+		__field(void*, workload)
 		__dynamic_array(u32, raw_cmd, cmd_len)
+		__array(char, cmd_name, GVT_CMD_STR_LEN)
 	),
 
 	TP_fast_assign(
@@ -244,17 +250,25 @@ TRACE_EVENT(gvt_command,
 		__entry->ring_id = ring_id;
 		__entry->ip_gma = ip_gma;
 		__entry->buf_type = buf_type;
+		__entry->buf_addr_type = buf_addr_type;
 		__entry->cmd_len = cmd_len;
+		__entry->workload = workload;
+		snprintf(__entry->cmd_name, GVT_CMD_STR_LEN, "%s", cmd_name);
 		memcpy(__get_dynamic_array(raw_cmd), cmd_va, cmd_len * sizeof(*cmd_va));
 	),
 
 
-	TP_printk("vgpu%d ring %d: buf_type %u, ip_gma %08x, raw cmd %s",
+	TP_printk("vgpu%d ring %d: address_type %u, buf_type %u, ip_gma %08x,cmd (name=%s,len=%u,raw cmd=%s), workload=%p\n",
 		__entry->vgpu_id,
 		__entry->ring_id,
+		__entry->buf_addr_type,
 		__entry->buf_type,
 		__entry->ip_gma,
-		__print_array(__get_dynamic_array(raw_cmd), __entry->cmd_len, 4))
+		__entry->cmd_name,
+		__entry->cmd_len,
+		__print_array(__get_dynamic_array(raw_cmd),
+			__entry->cmd_len, 4),
+		__entry->workload)
 );
 
 #define GVT_TEMP_STR_LEN 10

drivers/gpu/drm/i915/i915_debugfs.c

@@ -377,16 +377,19 @@ static void print_batch_pool_stats(struct seq_file *m,
 	print_file_stats(m, "[k]batch pool", stats);
 }
 
-static int per_file_ctx_stats(int id, void *ptr, void *data)
+static int per_file_ctx_stats(int idx, void *ptr, void *data)
 {
 	struct i915_gem_context *ctx = ptr;
-	int n;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, ctx->i915, id) {
+		struct intel_context *ce = to_intel_context(ctx, engine);
 
-	for (n = 0; n < ARRAY_SIZE(ctx->engine); n++) {
-		if (ctx->engine[n].state)
-			per_file_stats(0, ctx->engine[n].state->obj, data);
-		if (ctx->engine[n].ring)
-			per_file_stats(0, ctx->engine[n].ring->vma->obj, data);
+		if (ce->state)
+			per_file_stats(0, ce->state->obj, data);
+		if (ce->ring)
+			per_file_stats(0, ce->ring->vma->obj, data);
 	}
 
 	return 0;
@@ -1340,10 +1343,9 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 		struct rb_node *rb;
 
 		seq_printf(m, "%s:\n", engine->name);
-		seq_printf(m, "\tseqno = %x [current %x, last %x], inflight %d\n",
+		seq_printf(m, "\tseqno = %x [current %x, last %x]\n",
 			   engine->hangcheck.seqno, seqno[id],
-			   intel_engine_last_submit(engine),
-			   engine->timeline->inflight_seqnos);
+			   intel_engine_last_submit(engine));
 		seq_printf(m, "\twaiters? %s, fake irq active? %s, stalled? %s\n",
 			   yesno(intel_engine_has_waiter(engine)),
 			   yesno(test_bit(engine->id,
@@ -1960,7 +1962,8 @@ static int i915_context_status(struct seq_file *m, void *unused)
 		seq_putc(m, '\n');
 
 		for_each_engine(engine, dev_priv, id) {
-			struct intel_context *ce = &ctx->engine[engine->id];
+			struct intel_context *ce =
+				to_intel_context(ctx, engine);
 
 			seq_printf(m, "%s: ", engine->name);
 			if (ce->state)
@@ -2603,6 +2606,26 @@ static const char *psr2_live_status(u32 val)
 	return "unknown";
 }
 
+static const char *psr_sink_status(u8 val)
+{
+	static const char * const sink_status[] = {
+		"inactive",
+		"transition to active, capture and display",
+		"active, display from RFB",
+		"active, capture and display on sink device timings",
+		"transition to inactive, capture and display, timing re-sync",
+		"reserved",
+		"reserved",
+		"sink internal error"
+	};
+
+	val &= DP_PSR_SINK_STATE_MASK;
+	if (val < ARRAY_SIZE(sink_status))
+		return sink_status[val];
+
+	return "unknown";
+}
+
 static int i915_edp_psr_status(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -2684,12 +2707,61 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 		seq_printf(m, "EDP_PSR2_STATUS: %x [%s]\n",
 			   psr2, psr2_live_status(psr2));
 	}
+
+	if (dev_priv->psr.enabled) {
+		struct drm_dp_aux *aux = &dev_priv->psr.enabled->aux;
+		u8 val;
+
+		if (drm_dp_dpcd_readb(aux, DP_PSR_STATUS, &val) == 1)
+			seq_printf(m, "Sink PSR status: 0x%x [%s]\n", val,
+				   psr_sink_status(val));
+	}
 	mutex_unlock(&dev_priv->psr.lock);
 
+	if (READ_ONCE(dev_priv->psr.debug)) {
+		seq_printf(m, "Last attempted entry at: %lld\n",
+			   dev_priv->psr.last_entry_attempt);
+		seq_printf(m, "Last exit at: %lld\n",
+			   dev_priv->psr.last_exit);
+	}
+
+	intel_runtime_pm_put(dev_priv);
+	return 0;
+}
+
+static int
+i915_edp_psr_debug_set(void *data, u64 val)
+{
+	struct drm_i915_private *dev_priv = data;
+
+	if (!CAN_PSR(dev_priv))
+		return -ENODEV;
+
+	DRM_DEBUG_KMS("PSR debug %s\n", enableddisabled(val));
+
+	intel_runtime_pm_get(dev_priv);
+	intel_psr_irq_control(dev_priv, !!val);
 	intel_runtime_pm_put(dev_priv);
+
 	return 0;
 }
 
+static int
+i915_edp_psr_debug_get(void *data, u64 *val)
+{
+	struct drm_i915_private *dev_priv = data;
+
+	if (!CAN_PSR(dev_priv))
+		return -ENODEV;
+
+	*val = READ_ONCE(dev_priv->psr.debug);
+	return 0;
+}
+
+DEFINE_SIMPLE_ATTRIBUTE(i915_edp_psr_debug_fops,
+			i915_edp_psr_debug_get, i915_edp_psr_debug_set,
+			"%llu\n");
+
 static int i915_sink_crc(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -3296,6 +3368,28 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 		seq_printf(m, " fp0:     0x%08x\n", pll->state.hw_state.fp0);
 		seq_printf(m, " fp1:     0x%08x\n", pll->state.hw_state.fp1);
 		seq_printf(m, " wrpll:   0x%08x\n", pll->state.hw_state.wrpll);
+		seq_printf(m, " cfgcr0:  0x%08x\n", pll->state.hw_state.cfgcr0);
+		seq_printf(m, " cfgcr1:  0x%08x\n", pll->state.hw_state.cfgcr1);
+		seq_printf(m, " mg_refclkin_ctl:        0x%08x\n",
+			   pll->state.hw_state.mg_refclkin_ctl);
+		seq_printf(m, " mg_clktop2_coreclkctl1: 0x%08x\n",
+			   pll->state.hw_state.mg_clktop2_coreclkctl1);
+		seq_printf(m, " mg_clktop2_hsclkctl:    0x%08x\n",
+			   pll->state.hw_state.mg_clktop2_hsclkctl);
+		seq_printf(m, " mg_pll_div0:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_div0);
+		seq_printf(m, " mg_pll_div1:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_div1);
+		seq_printf(m, " mg_pll_lf:    0x%08x\n",
+			   pll->state.hw_state.mg_pll_lf);
+		seq_printf(m, " mg_pll_frac_lock: 0x%08x\n",
+			   pll->state.hw_state.mg_pll_frac_lock);
+		seq_printf(m, " mg_pll_ssc:   0x%08x\n",
+			   pll->state.hw_state.mg_pll_ssc);
+		seq_printf(m, " mg_pll_bias:  0x%08x\n",
+			   pll->state.hw_state.mg_pll_bias);
+		seq_printf(m, " mg_pll_tdc_coldst_bias: 0x%08x\n",
+			   pll->state.hw_state.mg_pll_tdc_coldst_bias);
 	}
 	drm_modeset_unlock_all(dev);
 
@@ -3304,24 +3398,13 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 
 static int i915_wa_registers(struct seq_file *m, void *unused)
 {
-	int i;
-	int ret;
-	struct intel_engine_cs *engine;
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
 	struct i915_workarounds *workarounds = &dev_priv->workarounds;
-	enum intel_engine_id id;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
+	int i;
 
 	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "Workarounds applied: %d\n", workarounds->count);
-	for_each_engine(engine, dev_priv, id)
-		seq_printf(m, "HW whitelist count for %s: %d\n",
-			   engine->name, workarounds->hw_whitelist_count[id]);
 	for (i = 0; i < workarounds->count; ++i) {
 		i915_reg_t addr;
 		u32 mask, value, read;
@@ -3337,7 +3420,6 @@ static int i915_wa_registers(struct seq_file *m, void *unused)
 	}
 
 	intel_runtime_pm_put(dev_priv);
-	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
@@ -4176,119 +4258,6 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_drop_caches_fops,
 			i915_drop_caches_get, i915_drop_caches_set,
 			"0x%08llx\n");
 
-static int
-i915_max_freq_get(void *data, u64 *val)
-{
-	struct drm_i915_private *dev_priv = data;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	*val = intel_gpu_freq(dev_priv, dev_priv->gt_pm.rps.max_freq_softlimit);
-	return 0;
-}
-
-static int
-i915_max_freq_set(void *data, u64 val)
-{
-	struct drm_i915_private *dev_priv = data;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 hw_max, hw_min;
-	int ret;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	DRM_DEBUG_DRIVER("Manually setting max freq to %llu\n", val);
-
-	ret = mutex_lock_interruptible(&dev_priv->pcu_lock);
-	if (ret)
-		return ret;
-
-	/*
-	 * Turbo will still be enabled, but won't go above the set value.
-	 */
-	val = intel_freq_opcode(dev_priv, val);
-
-	hw_max = rps->max_freq;
-	hw_min = rps->min_freq;
-
-	if (val < hw_min || val > hw_max || val < rps->min_freq_softlimit) {
-		mutex_unlock(&dev_priv->pcu_lock);
-		return -EINVAL;
-	}
-
-	rps->max_freq_softlimit = val;
-
-	if (intel_set_rps(dev_priv, val))
-		DRM_DEBUG_DRIVER("failed to update RPS to new softlimit\n");
-
-	mutex_unlock(&dev_priv->pcu_lock);
-
-	return 0;
-}
-
-DEFINE_SIMPLE_ATTRIBUTE(i915_max_freq_fops,
-			i915_max_freq_get, i915_max_freq_set,
-			"%llu\n");
-
-static int
-i915_min_freq_get(void *data, u64 *val)
-{
-	struct drm_i915_private *dev_priv = data;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	*val = intel_gpu_freq(dev_priv, dev_priv->gt_pm.rps.min_freq_softlimit);
-	return 0;
-}
-
-static int
-i915_min_freq_set(void *data, u64 val)
-{
-	struct drm_i915_private *dev_priv = data;
-	struct intel_rps *rps = &dev_priv->gt_pm.rps;
-	u32 hw_max, hw_min;
-	int ret;
-
-	if (INTEL_GEN(dev_priv) < 6)
-		return -ENODEV;
-
-	DRM_DEBUG_DRIVER("Manually setting min freq to %llu\n", val);
-
-	ret = mutex_lock_interruptible(&dev_priv->pcu_lock);
-	if (ret)
-		return ret;
-
-	/*
-	 * Turbo will still be enabled, but won't go below the set value.
-	 */
-	val = intel_freq_opcode(dev_priv, val);
-
-	hw_max = rps->max_freq;
-	hw_min = rps->min_freq;
-
-	if (val < hw_min ||
-	    val > hw_max || val > rps->max_freq_softlimit) {
-		mutex_unlock(&dev_priv->pcu_lock);
-		return -EINVAL;
-	}
-
-	rps->min_freq_softlimit = val;
-
-	if (intel_set_rps(dev_priv, val))
-		DRM_DEBUG_DRIVER("failed to update RPS to new softlimit\n");
-
-	mutex_unlock(&dev_priv->pcu_lock);
-
-	return 0;
-}
-
-DEFINE_SIMPLE_ATTRIBUTE(i915_min_freq_fops,
-			i915_min_freq_get, i915_min_freq_set,
-			"%llu\n");
-
 static int
 i915_cache_sharing_get(void *data, u64 *val)
 {
@@ -4850,8 +4819,6 @@ static const struct i915_debugfs_files {
 	const struct file_operations *fops;
 } i915_debugfs_files[] = {
 	{"i915_wedged", &i915_wedged_fops},
-	{"i915_max_freq", &i915_max_freq_fops},
-	{"i915_min_freq", &i915_min_freq_fops},
 	{"i915_cache_sharing", &i915_cache_sharing_fops},
 	{"i915_ring_missed_irq", &i915_ring_missed_irq_fops},
 	{"i915_ring_test_irq", &i915_ring_test_irq_fops},
@@ -4874,7 +4841,8 @@ static const struct i915_debugfs_files {
 	{"i915_guc_log_relay", &i915_guc_log_relay_fops},
 	{"i915_hpd_storm_ctl", &i915_hpd_storm_ctl_fops},
 	{"i915_ipc_status", &i915_ipc_status_fops},
-	{"i915_drrs_ctl", &i915_drrs_ctl_fops}
+	{"i915_drrs_ctl", &i915_drrs_ctl_fops},
+	{"i915_edp_psr_debug", &i915_edp_psr_debug_fops}
 };
 
 int i915_debugfs_register(struct drm_i915_private *dev_priv)

drivers/gpu/drm/i915/i915_drv.c

@@ -101,7 +101,13 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level,
 		   __builtin_return_address(0), &vaf);
 
 	if (is_error && !shown_bug_once) {
-		dev_notice(kdev, "%s", FDO_BUG_MSG);
+		/*
+		 * Ask the user to file a bug report for the error, except
+		 * if they may have caused the bug by fiddling with unsafe
+		 * module parameters.
+		 */
+		if (!test_taint(TAINT_USER))
+			dev_notice(kdev, "%s", FDO_BUG_MSG);
 		shown_bug_once = true;
 	}
 
@@ -2468,10 +2474,13 @@ static void vlv_wait_for_gt_wells(struct drm_i915_private *dev_priv,
 	/*
 	 * RC6 transitioning can be delayed up to 2 msec (see
 	 * valleyview_enable_rps), use 3 msec for safety.
+	 *
+	 * This can fail to turn off the rc6 if the GPU is stuck after a failed
+	 * reset and we are trying to force the machine to sleep.
 	 */
 	if (vlv_wait_for_pw_status(dev_priv, mask, val))
-		DRM_ERROR("timeout waiting for GT wells to go %s\n",
-			  onoff(wait_for_on));
+		DRM_DEBUG_DRIVER("timeout waiting for GT wells to go %s\n",
+				 onoff(wait_for_on));
 }
 
 static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)

drivers/gpu/drm/i915/i915_drv.h

@@ -72,9 +72,10 @@
 #include "i915_gem_fence_reg.h"
 #include "i915_gem_object.h"
 #include "i915_gem_gtt.h"
-#include "i915_gem_timeline.h"
 #include "i915_gpu_error.h"
 #include "i915_request.h"
+#include "i915_scheduler.h"
+#include "i915_timeline.h"
 #include "i915_vma.h"
 
 #include "intel_gvt.h"
@@ -84,8 +85,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20180413"
-#define DRIVER_TIMESTAMP	1523611258
+#define DRIVER_DATE		"20180514"
+#define DRIVER_TIMESTAMP	1526300884
 
 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and
  * WARN_ON()) for hw state sanity checks to check for unexpected conditions
@@ -609,6 +610,9 @@ struct i915_psr {
 	bool has_hw_tracking;
 	bool psr2_enabled;
 	u8 sink_sync_latency;
+	bool debug;
+	ktime_t last_entry_attempt;
+	ktime_t last_exit;
 
 	void (*enable_source)(struct intel_dp *,
 			      const struct intel_crtc_state *);
@@ -1069,6 +1073,7 @@ struct intel_vbt_data {
 	} edp;
 
 	struct {
+		bool enable;
 		bool full_link;
 		bool require_aux_wakeup;
 		int idle_frames;
@@ -1185,6 +1190,7 @@ struct skl_ddb_allocation {
 	/* packed/y */
 	struct skl_ddb_entry plane[I915_MAX_PIPES][I915_MAX_PLANES];
 	struct skl_ddb_entry uv_plane[I915_MAX_PIPES][I915_MAX_PLANES];
+	u8 enabled_slices; /* GEN11 has configurable 2 slices */
 };
 
 struct skl_ddb_values {
@@ -1297,7 +1303,6 @@ struct i915_wa_reg {
 struct i915_workarounds {
 	struct i915_wa_reg reg[I915_MAX_WA_REGS];
 	u32 count;
-	u32 hw_whitelist_count[I915_NUM_ENGINES];
 };
 
 struct i915_virtual_gpu {
@@ -2056,8 +2061,11 @@ struct drm_i915_private {
 		void (*cleanup_engine)(struct intel_engine_cs *engine);
 
 		struct list_head timelines;
-		struct i915_gem_timeline global_timeline;
+
+		struct list_head active_rings;
+		struct list_head closed_vma;
 		u32 active_requests;
+		u32 request_serial;
 
 		/**
 		 * Is the GPU currently considered idle, or busy executing
@@ -2462,6 +2470,15 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_CNL_REVID(p, since, until) \
 	(IS_CANNONLAKE(p) && IS_REVID(p, since, until))
 
+#define ICL_REVID_A0		0x0
+#define ICL_REVID_A2		0x1
+#define ICL_REVID_B0		0x3
+#define ICL_REVID_B2		0x4
+#define ICL_REVID_C0		0x5
+
+#define IS_ICL_REVID(p, since, until) \
+	(IS_ICELAKE(p) && IS_REVID(p, since, until))
+
 /*
  * The genX designation typically refers to the render engine, so render
  * capability related checks should use IS_GEN, while display and other checks
@@ -3159,7 +3176,7 @@ int i915_gem_object_wait(struct drm_i915_gem_object *obj,
 			 struct intel_rps_client *rps);
 int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 				  unsigned int flags,
-				  int priority);
+				  const struct i915_sched_attr *attr);
 #define I915_PRIORITY_DISPLAY I915_PRIORITY_MAX
 
 int __must_check
@@ -3228,16 +3245,6 @@ i915_gem_context_lookup(struct drm_i915_file_private *file_priv, u32 id)
 	return ctx;
 }
 
-static inline struct intel_timeline *
-i915_gem_context_lookup_timeline(struct i915_gem_context *ctx,
-				 struct intel_engine_cs *engine)
-{
-	struct i915_address_space *vm;
-
-	vm = ctx->ppgtt ? &ctx->ppgtt->base : &ctx->i915->ggtt.base;
-	return &vm->timeline.engine[engine->id];
-}
-
 int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file);
 int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,

drivers/gpu/drm/i915/i915_gem.c

@@ -141,6 +141,7 @@ static u32 __i915_gem_park(struct drm_i915_private *i915)
 {
 	lockdep_assert_held(&i915->drm.struct_mutex);
 	GEM_BUG_ON(i915->gt.active_requests);
+	GEM_BUG_ON(!list_empty(&i915->gt.active_rings));
 
 	if (!i915->gt.awake)
 		return I915_EPOCH_INVALID;
@@ -161,9 +162,10 @@ static u32 __i915_gem_park(struct drm_i915_private *i915)
 	synchronize_irq(i915->drm.irq);
 
 	intel_engines_park(i915);
-	i915_gem_timelines_park(i915);
+	i915_timelines_park(i915);
 
 	i915_pmu_gt_parked(i915);
+	i915_vma_parked(i915);
 
 	i915->gt.awake = false;
 
@@ -564,7 +566,8 @@ i915_gem_object_wait_reservation(struct reservation_object *resv,
 	return timeout;
 }
 
-static void __fence_set_priority(struct dma_fence *fence, int prio)
+static void __fence_set_priority(struct dma_fence *fence,
+				 const struct i915_sched_attr *attr)
 {
 	struct i915_request *rq;
 	struct intel_engine_cs *engine;
@@ -575,13 +578,16 @@ static void __fence_set_priority(struct dma_fence *fence, int prio)
 	rq = to_request(fence);
 	engine = rq->engine;
 
-	rcu_read_lock();
+	local_bh_disable();
+	rcu_read_lock(); /* RCU serialisation for set-wedged protection */
 	if (engine->schedule)
-		engine->schedule(rq, prio);
+		engine->schedule(rq, attr);
 	rcu_read_unlock();
+	local_bh_enable(); /* kick the tasklets if queues were reprioritised */
 }
 
-static void fence_set_priority(struct dma_fence *fence, int prio)
+static void fence_set_priority(struct dma_fence *fence,
+			       const struct i915_sched_attr *attr)
 {
 	/* Recurse once into a fence-array */
 	if (dma_fence_is_array(fence)) {
@@ -589,16 +595,16 @@ static void fence_set_priority(struct dma_fence *fence, int prio)
 		int i;
 
 		for (i = 0; i < array->num_fences; i++)
-			__fence_set_priority(array->fences[i], prio);
+			__fence_set_priority(array->fences[i], attr);
 	} else {
-		__fence_set_priority(fence, prio);
+		__fence_set_priority(fence, attr);
 	}
 }
 
 int
 i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 			      unsigned int flags,
-			      int prio)
+			      const struct i915_sched_attr *attr)
 {
 	struct dma_fence *excl;
 
@@ -613,7 +619,7 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 			return ret;
 
 		for (i = 0; i < count; i++) {
-			fence_set_priority(shared[i], prio);
+			fence_set_priority(shared[i], attr);
 			dma_fence_put(shared[i]);
 		}
 
@@ -623,7 +629,7 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 	}
 
 	if (excl) {
-		fence_set_priority(excl, prio);
+		fence_set_priority(excl, attr);
 		dma_fence_put(excl);
 	}
 	return 0;
@@ -2974,8 +2980,8 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 	 * extra delay for a recent interrupt is pointless. Hence, we do
 	 * not need an engine->irq_seqno_barrier() before the seqno reads.
 	 */
-	spin_lock_irqsave(&engine->timeline->lock, flags);
-	list_for_each_entry(request, &engine->timeline->requests, link) {
+	spin_lock_irqsave(&engine->timeline.lock, flags);
+	list_for_each_entry(request, &engine->timeline.requests, link) {
 		if (__i915_request_completed(request, request->global_seqno))
 			continue;
 
@@ -2986,7 +2992,7 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 		active = request;
 		break;
 	}
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 
 	return active;
 }
@@ -3107,15 +3113,15 @@ static void engine_skip_context(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct i915_gem_context *hung_ctx = request->ctx;
-	struct intel_timeline *timeline;
+	struct i915_timeline *timeline = request->timeline;
 	unsigned long flags;
 
-	timeline = i915_gem_context_lookup_timeline(hung_ctx, engine);
+	GEM_BUG_ON(timeline == &engine->timeline);
 
-	spin_lock_irqsave(&engine->timeline->lock, flags);
-	spin_lock(&timeline->lock);
+	spin_lock_irqsave(&engine->timeline.lock, flags);
+	spin_lock_nested(&timeline->lock, SINGLE_DEPTH_NESTING);
 
-	list_for_each_entry_continue(request, &engine->timeline->requests, link)
+	list_for_each_entry_continue(request, &engine->timeline.requests, link)
 		if (request->ctx == hung_ctx)
 			skip_request(request);
 
@@ -3123,7 +3129,7 @@ static void engine_skip_context(struct i915_request *request)
 		skip_request(request);
 
 	spin_unlock(&timeline->lock);
-	spin_unlock_irqrestore(&engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&engine->timeline.lock, flags);
 }
 
 /* Returns the request if it was guilty of the hang */
@@ -3180,11 +3186,11 @@ i915_gem_reset_request(struct intel_engine_cs *engine,
 			dma_fence_set_error(&request->fence, -EAGAIN);
 
 			/* Rewind the engine to replay the incomplete rq */
-			spin_lock_irq(&engine->timeline->lock);
+			spin_lock_irq(&engine->timeline.lock);
 			request = list_prev_entry(request, link);
-			if (&request->link == &engine->timeline->requests)
+			if (&request->link == &engine->timeline.requests)
 				request = NULL;
-			spin_unlock_irq(&engine->timeline->lock);
+			spin_unlock_irq(&engine->timeline.lock);
 		}
 	}
 
@@ -3232,7 +3238,7 @@ void i915_gem_reset(struct drm_i915_private *dev_priv,
 				      stalled_mask & ENGINE_MASK(id));
 		ctx = fetch_and_zero(&engine->last_retired_context);
 		if (ctx)
-			engine->context_unpin(engine, ctx);
+			intel_context_unpin(ctx, engine);
 
 		/*
 		 * Ostensibily, we always want a context loaded for powersaving,
@@ -3297,10 +3303,10 @@ static void nop_complete_submit_request(struct i915_request *request)
 		  request->fence.context, request->fence.seqno);
 	dma_fence_set_error(&request->fence, -EIO);
 
-	spin_lock_irqsave(&request->engine->timeline->lock, flags);
+	spin_lock_irqsave(&request->engine->timeline.lock, flags);
 	__i915_request_submit(request);
 	intel_engine_init_global_seqno(request->engine, request->global_seqno);
-	spin_unlock_irqrestore(&request->engine->timeline->lock, flags);
+	spin_unlock_irqrestore(&request->engine->timeline.lock, flags);
 }
 
 void i915_gem_set_wedged(struct drm_i915_private *i915)
@@ -3310,7 +3316,7 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 
 	GEM_TRACE("start\n");
 
-	if (drm_debug & DRM_UT_DRIVER) {
+	if (GEM_SHOW_DEBUG()) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
 		for_each_engine(engine, i915, id)
@@ -3369,10 +3375,10 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 		 * (lockless) lookup doesn't try and wait upon the request as we
 		 * reset it.
 		 */
-		spin_lock_irqsave(&engine->timeline->lock, flags);
+		spin_lock_irqsave(&engine->timeline.lock, flags);
 		intel_engine_init_global_seqno(engine,
 					       intel_engine_last_submit(engine));
-		spin_unlock_irqrestore(&engine->timeline->lock, flags);
+		spin_unlock_irqrestore(&engine->timeline.lock, flags);
 
 		i915_gem_reset_finish_engine(engine);
 	}
@@ -3384,8 +3390,7 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 
 bool i915_gem_unset_wedged(struct drm_i915_private *i915)
 {
-	struct i915_gem_timeline *tl;
-	int i;
+	struct i915_timeline *tl;
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 	if (!test_bit(I915_WEDGED, &i915->gpu_error.flags))
@@ -3404,29 +3409,27 @@ bool i915_gem_unset_wedged(struct drm_i915_private *i915)
 	 * No more can be submitted until we reset the wedged bit.
 	 */
 	list_for_each_entry(tl, &i915->gt.timelines, link) {
-		for (i = 0; i < ARRAY_SIZE(tl->engine); i++) {
-			struct i915_request *rq;
+		struct i915_request *rq;
 
-			rq = i915_gem_active_peek(&tl->engine[i].last_request,
-						  &i915->drm.struct_mutex);
-			if (!rq)
-				continue;
+		rq = i915_gem_active_peek(&tl->last_request,
+					  &i915->drm.struct_mutex);
+		if (!rq)
+			continue;
 
-			/*
-			 * We can't use our normal waiter as we want to
-			 * avoid recursively trying to handle the current
-			 * reset. The basic dma_fence_default_wait() installs
-			 * a callback for dma_fence_signal(), which is
-			 * triggered by our nop handler (indirectly, the
-			 * callback enables the signaler thread which is
-			 * woken by the nop_submit_request() advancing the seqno
-			 * and when the seqno passes the fence, the signaler
-			 * then signals the fence waking us up).
-			 */
-			if (dma_fence_default_wait(&rq->fence, true,
-						   MAX_SCHEDULE_TIMEOUT) < 0)
-				return false;
-		}
+		/*
+		 * We can't use our normal waiter as we want to
+		 * avoid recursively trying to handle the current
+		 * reset. The basic dma_fence_default_wait() installs
+		 * a callback for dma_fence_signal(), which is
+		 * triggered by our nop handler (indirectly, the
+		 * callback enables the signaler thread which is
+		 * woken by the nop_submit_request() advancing the seqno
+		 * and when the seqno passes the fence, the signaler
+		 * then signals the fence waking us up).
+		 */
+		if (dma_fence_default_wait(&rq->fence, true,
+					   MAX_SCHEDULE_TIMEOUT) < 0)
+			return false;
 	}
 	i915_retire_requests(i915);
 	GEM_BUG_ON(i915->gt.active_requests);
@@ -3731,17 +3734,9 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	return ret;
 }
 
-static int wait_for_timeline(struct i915_gem_timeline *tl, unsigned int flags)
+static int wait_for_timeline(struct i915_timeline *tl, unsigned int flags)
 {
-	int ret, i;
-
-	for (i = 0; i < ARRAY_SIZE(tl->engine); i++) {
-		ret = i915_gem_active_wait(&tl->engine[i].last_request, flags);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
+	return i915_gem_active_wait(&tl->last_request, flags);
 }
 
 static int wait_for_engines(struct drm_i915_private *i915)
@@ -3759,30 +3754,37 @@ static int wait_for_engines(struct drm_i915_private *i915)
 
 int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags)
 {
-	int ret;
-
 	/* If the device is asleep, we have no requests outstanding */
 	if (!READ_ONCE(i915->gt.awake))
 		return 0;
 
 	if (flags & I915_WAIT_LOCKED) {
-		struct i915_gem_timeline *tl;
+		struct i915_timeline *tl;
+		int err;
 
 		lockdep_assert_held(&i915->drm.struct_mutex);
 
 		list_for_each_entry(tl, &i915->gt.timelines, link) {
-			ret = wait_for_timeline(tl, flags);
-			if (ret)
-				return ret;
+			err = wait_for_timeline(tl, flags);
+			if (err)
+				return err;
 		}
 		i915_retire_requests(i915);
 
-		ret = wait_for_engines(i915);
+		return wait_for_engines(i915);
 	} else {
-		ret = wait_for_timeline(&i915->gt.global_timeline, flags);
-	}
+		struct intel_engine_cs *engine;
+		enum intel_engine_id id;
+		int err;
 
-	return ret;
+		for_each_engine(engine, i915, id) {
+			err = wait_for_timeline(&engine->timeline, flags);
+			if (err)
+				return err;
+		}
+
+		return 0;
+	}
 }
 
 static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
@@ -4796,7 +4798,7 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 					 &obj->vma_list, obj_link) {
 			GEM_BUG_ON(i915_vma_is_active(vma));
 			vma->flags &= ~I915_VMA_PIN_MASK;
-			i915_vma_close(vma);
+			i915_vma_destroy(vma);
 		}
 		GEM_BUG_ON(!list_empty(&obj->vma_list));
 		GEM_BUG_ON(!RB_EMPTY_ROOT(&obj->vma_tree));
@@ -4951,7 +4953,7 @@ static void assert_kernel_context_is_current(struct drm_i915_private *i915)
 	enum intel_engine_id id;
 
 	for_each_engine(engine, i915, id) {
-		GEM_BUG_ON(__i915_gem_active_peek(&engine->timeline->last_request));
+		GEM_BUG_ON(__i915_gem_active_peek(&engine->timeline.last_request));
 		GEM_BUG_ON(engine->last_retired_context != kernel_context);
 	}
 }
@@ -5289,7 +5291,7 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 	for_each_engine(engine, i915, id) {
 		struct i915_vma *state;
 
-		state = ctx->engine[id].state;
+		state = to_intel_context(ctx, engine)->state;
 		if (!state)
 			continue;
 
@@ -5597,12 +5599,9 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 	if (!dev_priv->priorities)
 		goto err_dependencies;
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
-	err = i915_gem_timeline_init__global(dev_priv);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-	if (err)
-		goto err_priorities;
+	INIT_LIST_HEAD(&dev_priv->gt.active_rings);
+	INIT_LIST_HEAD(&dev_priv->gt.closed_vma);
 
 	i915_gem_init__mm(dev_priv);
 
@@ -5623,8 +5622,6 @@ int i915_gem_init_early(struct drm_i915_private *dev_priv)
 
 	return 0;
 
-err_priorities:
-	kmem_cache_destroy(dev_priv->priorities);
 err_dependencies:
 	kmem_cache_destroy(dev_priv->dependencies);
 err_requests:
@@ -5645,11 +5642,7 @@ void i915_gem_cleanup_early(struct drm_i915_private *dev_priv)
 	GEM_BUG_ON(!llist_empty(&dev_priv->mm.free_list));
 	GEM_BUG_ON(atomic_read(&dev_priv->mm.free_count));
 	WARN_ON(dev_priv->mm.object_count);
-
-	mutex_lock(&dev_priv->drm.struct_mutex);
-	i915_gem_timeline_fini(&dev_priv->gt.global_timeline);
 	WARN_ON(!list_empty(&dev_priv->gt.timelines));
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	kmem_cache_destroy(dev_priv->priorities);
 	kmem_cache_destroy(dev_priv->dependencies);

drivers/gpu/drm/i915/i915_gem.h

@@ -30,6 +30,9 @@
 struct drm_i915_private;
 
 #ifdef CONFIG_DRM_I915_DEBUG_GEM
+
+#define GEM_SHOW_DEBUG() (drm_debug & DRM_UT_DRIVER)
+
 #define GEM_BUG_ON(condition) do { if (unlikely((condition))) {	\
 		pr_err("%s:%d GEM_BUG_ON(%s)\n", \
 		       __func__, __LINE__, __stringify(condition)); \
@@ -45,6 +48,9 @@ struct drm_i915_private;
 #define GEM_DEBUG_BUG_ON(expr) GEM_BUG_ON(expr)
 
 #else
+
+#define GEM_SHOW_DEBUG() (0)
+
 #define GEM_BUG_ON(expr) BUILD_BUG_ON_INVALID(expr)
 #define GEM_WARN_ON(expr) (BUILD_BUG_ON_INVALID(expr), 0)
 

drivers/gpu/drm/i915/i915_gem_context.c

@@ -117,15 +117,15 @@ static void lut_close(struct i915_gem_context *ctx)
 
 static void i915_gem_context_free(struct i915_gem_context *ctx)
 {
-	int i;
+	unsigned int n;
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 	GEM_BUG_ON(!i915_gem_context_is_closed(ctx));
 
 	i915_ppgtt_put(ctx->ppgtt);
 
-	for (i = 0; i < I915_NUM_ENGINES; i++) {
-		struct intel_context *ce = &ctx->engine[i];
+	for (n = 0; n < ARRAY_SIZE(ctx->__engine); n++) {
+		struct intel_context *ce = &ctx->__engine[n];
 
 		if (!ce->state)
 			continue;
@@ -281,7 +281,7 @@ __create_hw_context(struct drm_i915_private *dev_priv,
 	kref_init(&ctx->ref);
 	list_add_tail(&ctx->link, &dev_priv->contexts.list);
 	ctx->i915 = dev_priv;
-	ctx->priority = I915_PRIORITY_NORMAL;
+	ctx->sched.priority = I915_PRIORITY_NORMAL;
 
 	INIT_RADIX_TREE(&ctx->handles_vma, GFP_KERNEL);
 	INIT_LIST_HEAD(&ctx->handles_list);
@@ -431,7 +431,7 @@ i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio)
 		return ctx;
 
 	i915_gem_context_clear_bannable(ctx);
-	ctx->priority = prio;
+	ctx->sched.priority = prio;
 	ctx->ring_size = PAGE_SIZE;
 
 	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
@@ -521,7 +521,7 @@ void i915_gem_contexts_lost(struct drm_i915_private *dev_priv)
 		if (!engine->last_retired_context)
 			continue;
 
-		engine->context_unpin(engine, engine->last_retired_context);
+		intel_context_unpin(engine->last_retired_context, engine);
 		engine->last_retired_context = NULL;
 	}
 }
@@ -577,19 +577,29 @@ void i915_gem_context_close(struct drm_file *file)
 	idr_destroy(&file_priv->context_idr);
 }
 
-static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
+static struct i915_request *
+last_request_on_engine(struct i915_timeline *timeline,
+		       struct intel_engine_cs *engine)
 {
-	struct i915_gem_timeline *timeline;
+	struct i915_request *rq;
 
-	list_for_each_entry(timeline, &engine->i915->gt.timelines, link) {
-		struct intel_timeline *tl;
+	if (timeline == &engine->timeline)
+		return NULL;
 
-		if (timeline == &engine->i915->gt.global_timeline)
-			continue;
+	rq = i915_gem_active_raw(&timeline->last_request,
+				 &engine->i915->drm.struct_mutex);
+	if (rq && rq->engine == engine)
+		return rq;
+
+	return NULL;
+}
 
-		tl = &timeline->engine[engine->id];
-		if (i915_gem_active_peek(&tl->last_request,
-					 &engine->i915->drm.struct_mutex))
+static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
+{
+	struct i915_timeline *timeline;
+
+	list_for_each_entry(timeline, &engine->i915->gt.timelines, link) {
+		if (last_request_on_engine(timeline, engine))
 			return false;
 	}
 
@@ -599,7 +609,7 @@ static bool engine_has_idle_kernel_context(struct intel_engine_cs *engine)
 int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 {
 	struct intel_engine_cs *engine;
-	struct i915_gem_timeline *timeline;
+	struct i915_timeline *timeline;
 	enum intel_engine_id id;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
@@ -619,11 +629,8 @@ int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 		/* Queue this switch after all other activity */
 		list_for_each_entry(timeline, &dev_priv->gt.timelines, link) {
 			struct i915_request *prev;
-			struct intel_timeline *tl;
 
-			tl = &timeline->engine[engine->id];
-			prev = i915_gem_active_raw(&tl->last_request,
-						   &dev_priv->drm.struct_mutex);
+			prev = last_request_on_engine(timeline, engine);
 			if (prev)
 				i915_sw_fence_await_sw_fence_gfp(&rq->submit,
 								 &prev->submit,
@@ -753,7 +760,7 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data,
 		args->value = i915_gem_context_is_bannable(ctx);
 		break;
 	case I915_CONTEXT_PARAM_PRIORITY:
-		args->value = ctx->priority;
+		args->value = ctx->sched.priority;
 		break;
 	default:
 		ret = -EINVAL;
@@ -826,7 +833,7 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data,
 				 !capable(CAP_SYS_NICE))
 				ret = -EPERM;
 			else
-				ctx->priority = priority;
+				ctx->sched.priority = priority;
 		}
 		break;
 

drivers/gpu/drm/i915/i915_gem_context.h

@@ -137,18 +137,7 @@ struct i915_gem_context {
 	 */
 	u32 user_handle;
 
-	/**
-	 * @priority: execution and service priority
-	 *
-	 * All clients are equal, but some are more equal than others!
-	 *
-	 * Requests from a context with a greater (more positive) value of
-	 * @priority will be executed before those with a lower @priority
-	 * value, forming a simple QoS.
-	 *
-	 * The &drm_i915_private.kernel_context is assigned the lowest priority.
-	 */
-	int priority;
+	struct i915_sched_attr sched;
 
 	/** ggtt_offset_bias: placement restriction for context objects */
 	u32 ggtt_offset_bias;
@@ -160,7 +149,7 @@ struct i915_gem_context {
 		u32 *lrc_reg_state;
 		u64 lrc_desc;
 		int pin_count;
-	} engine[I915_NUM_ENGINES];
+	} __engine[I915_NUM_ENGINES];
 
 	/** ring_size: size for allocating the per-engine ring buffer */
 	u32 ring_size;
@@ -267,6 +256,34 @@ static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
 	return !ctx->file_priv;
 }
 
+static inline struct intel_context *
+to_intel_context(struct i915_gem_context *ctx,
+		 const struct intel_engine_cs *engine)
+{
+	return &ctx->__engine[engine->id];
+}
+
+static inline struct intel_ring *
+intel_context_pin(struct i915_gem_context *ctx, struct intel_engine_cs *engine)
+{
+	return engine->context_pin(engine, ctx);
+}
+
+static inline void __intel_context_pin(struct i915_gem_context *ctx,
+				       const struct intel_engine_cs *engine)
+{
+	struct intel_context *ce = to_intel_context(ctx, engine);
+
+	GEM_BUG_ON(!ce->pin_count);
+	ce->pin_count++;
+}
+
+static inline void intel_context_unpin(struct i915_gem_context *ctx,
+				       struct intel_engine_cs *engine)
+{
+	engine->context_unpin(engine, ctx);
+}
+
 /* i915_gem_context.c */
 int __must_check i915_gem_contexts_init(struct drm_i915_private *dev_priv);
 void i915_gem_contexts_lost(struct drm_i915_private *dev_priv);

drivers/gpu/drm/i915/i915_gem_execbuffer.c

@@ -762,7 +762,8 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 		}
 
 		/* transfer ref to ctx */
-		vma->open_count++;
+		if (!vma->open_count++)
+			i915_vma_reopen(vma);
 		list_add(&lut->obj_link, &obj->lut_list);
 		list_add(&lut->ctx_link, &eb->ctx->handles_list);
 		lut->ctx = eb->ctx;

drivers/gpu/drm/i915/i915_gem_gtt.c

@@ -110,7 +110,8 @@ i915_get_ggtt_vma_pages(struct i915_vma *vma);
 
 static void gen6_ggtt_invalidate(struct drm_i915_private *dev_priv)
 {
-	/* Note that as an uncached mmio write, this should flush the
+	/*
+	 * Note that as an uncached mmio write, this will flush the
 	 * WCB of the writes into the GGTT before it triggers the invalidate.
 	 */
 	I915_WRITE(GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
@@ -1161,6 +1162,27 @@ static void gen8_ppgtt_insert_huge_entries(struct i915_vma *vma,
 			vaddr[idx.pde] |= GEN8_PDE_IPS_64K;
 			kunmap_atomic(vaddr);
 			page_size = I915_GTT_PAGE_SIZE_64K;
+
+			/*
+			 * We write all 4K page entries, even when using 64K
+			 * pages. In order to verify that the HW isn't cheating
+			 * by using the 4K PTE instead of the 64K PTE, we want
+			 * to remove all the surplus entries. If the HW skipped
+			 * the 64K PTE, it will read/write into the scratch page
+			 * instead - which we detect as missing results during
+			 * selftests.
+			 */
+			if (I915_SELFTEST_ONLY(vma->vm->scrub_64K)) {
+				u16 i;
+
+				encode = pte_encode | vma->vm->scratch_page.daddr;
+				vaddr = kmap_atomic_px(pd->page_table[idx.pde]);
+
+				for (i = 1; i < index; i += 16)
+					memset64(vaddr + i, encode, 15);
+
+				kunmap_atomic(vaddr);
+			}
 		}
 
 		vma->page_sizes.gtt |= page_size;
@@ -2111,8 +2133,6 @@ static void i915_address_space_init(struct i915_address_space *vm,
 				    struct drm_i915_private *dev_priv,
 				    const char *name)
 {
-	i915_gem_timeline_init(dev_priv, &vm->timeline, name);
-
 	drm_mm_init(&vm->mm, 0, vm->total);
 	vm->mm.head_node.color = I915_COLOR_UNEVICTABLE;
 
@@ -2129,7 +2149,6 @@ static void i915_address_space_fini(struct i915_address_space *vm)
 	if (pagevec_count(&vm->free_pages))
 		vm_free_pages_release(vm, true);
 
-	i915_gem_timeline_fini(&vm->timeline);
 	drm_mm_takedown(&vm->mm);
 	list_del(&vm->global_link);
 }
@@ -2140,15 +2159,15 @@ static void gtt_write_workarounds(struct drm_i915_private *dev_priv)
 	 * called on driver load and after a GPU reset, so you can place
 	 * workarounds here even if they get overwritten by GPU reset.
 	 */
-	/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl */
+	/* WaIncreaseDefaultTLBEntries:chv,bdw,skl,bxt,kbl,glk,cfl,cnl,icl */
 	if (IS_BROADWELL(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_BDW);
 	else if (IS_CHERRYVIEW(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN8_L3_LRA_1_GPGPU_DEFAULT_VALUE_CHV);
-	else if (IS_GEN9_BC(dev_priv) || IS_GEN10(dev_priv))
-		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL);
 	else if (IS_GEN9_LP(dev_priv))
 		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_BXT);
+	else if (INTEL_GEN(dev_priv) >= 9)
+		I915_WRITE(GEN8_L3_LRA_1_GPGPU, GEN9_L3_LRA_1_GPGPU_DEFAULT_VALUE_SKL);
 
 	/*
 	 * To support 64K PTEs we need to first enable the use of the
@@ -2221,6 +2240,12 @@ i915_ppgtt_create(struct drm_i915_private *dev_priv,
 }
 
 void i915_ppgtt_close(struct i915_address_space *vm)
+{
+	GEM_BUG_ON(vm->closed);
+	vm->closed = true;
+}
+
+static void ppgtt_destroy_vma(struct i915_address_space *vm)
 {
 	struct list_head *phases[] = {
 		&vm->active_list,
@@ -2229,15 +2254,12 @@ void i915_ppgtt_close(struct i915_address_space *vm)
 		NULL,
 	}, **phase;
 
-	GEM_BUG_ON(vm->closed);
 	vm->closed = true;
-
 	for (phase = phases; *phase; phase++) {
 		struct i915_vma *vma, *vn;
 
 		list_for_each_entry_safe(vma, vn, *phase, vm_link)
-			if (!i915_vma_is_closed(vma))
-				i915_vma_close(vma);
+			i915_vma_destroy(vma);
 	}
 }
 
@@ -2248,7 +2270,8 @@ void i915_ppgtt_release(struct kref *kref)
 
 	trace_i915_ppgtt_release(&ppgtt->base);
 
-	/* vmas should already be unbound and destroyed */
+	ppgtt_destroy_vma(&ppgtt->base);
+
 	GEM_BUG_ON(!list_empty(&ppgtt->base.active_list));
 	GEM_BUG_ON(!list_empty(&ppgtt->base.inactive_list));
 	GEM_BUG_ON(!list_empty(&ppgtt->base.unbound_list));
@@ -2417,11 +2440,9 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 	for_each_sgt_dma(addr, sgt_iter, vma->pages)
 		gen8_set_pte(gtt_entries++, pte_encode | addr);
 
-	wmb();
-
-	/* This next bit makes the above posting read even more important. We
-	 * want to flush the TLBs only after we're certain all the PTE updates
-	 * have finished.
+	/*
+	 * We want to flush the TLBs only after we're certain all the PTE
+	 * updates have finished.
 	 */
 	ggtt->invalidate(vm->i915);
 }
@@ -2459,11 +2480,10 @@ static void gen6_ggtt_insert_entries(struct i915_address_space *vm,
 	dma_addr_t addr;
 	for_each_sgt_dma(addr, iter, vma->pages)
 		iowrite32(vm->pte_encode(addr, level, flags), &entries[i++]);
-	wmb();
 
-	/* This next bit makes the above posting read even more important. We
-	 * want to flush the TLBs only after we're certain all the PTE updates
-	 * have finished.
+	/*
+	 * We want to flush the TLBs only after we're certain all the PTE
+	 * updates have finished.
 	 */
 	ggtt->invalidate(vm->i915);
 }
@@ -3325,14 +3345,10 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 		DRM_ERROR("Can't set DMA mask/consistent mask (%d)\n", err);
 
 	pci_read_config_word(pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
-
-	if (INTEL_GEN(dev_priv) >= 9) {
-		size = gen8_get_total_gtt_size(snb_gmch_ctl);
-	} else if (IS_CHERRYVIEW(dev_priv)) {
+	if (IS_CHERRYVIEW(dev_priv))
 		size = chv_get_total_gtt_size(snb_gmch_ctl);
-	} else {
+	else
 		size = gen8_get_total_gtt_size(snb_gmch_ctl);
-	}
 
 	ggtt->base.total = (size / sizeof(gen8_pte_t)) << PAGE_SHIFT;
 	ggtt->base.cleanup = gen6_gmch_remove;

drivers/gpu/drm/i915/i915_gem_gtt.h

@@ -38,10 +38,9 @@
 #include <linux/mm.h>
 #include <linux/pagevec.h>
 
-#include "i915_gem_timeline.h"
-
 #include "i915_request.h"
 #include "i915_selftest.h"
+#include "i915_timeline.h"
 
 #define I915_GTT_PAGE_SIZE_4K BIT(12)
 #define I915_GTT_PAGE_SIZE_64K BIT(16)
@@ -257,7 +256,6 @@ struct i915_pml4 {
 
 struct i915_address_space {
 	struct drm_mm mm;
-	struct i915_gem_timeline timeline;
 	struct drm_i915_private *i915;
 	struct device *dma;
 	/* Every address space belongs to a struct file - except for the global
@@ -344,6 +342,7 @@ struct i915_address_space {
 	void (*clear_pages)(struct i915_vma *vma);
 
 	I915_SELFTEST_DECLARE(struct fault_attr fault_attr);
+	I915_SELFTEST_DECLARE(bool scrub_64K);
 };
 
 #define i915_is_ggtt(V) (!(V)->file)

drivers/gpu/drm/i915/i915_gem_stolen.c

@@ -51,6 +51,10 @@ int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *dev_priv,
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
 		return -ENODEV;
 
+	/* WaSkipStolenMemoryFirstPage:bdw+ */
+	if (INTEL_GEN(dev_priv) >= 8 && start < 4096)
+		start = 4096;
+
 	mutex_lock(&dev_priv->mm.stolen_lock);
 	ret = drm_mm_insert_node_in_range(&dev_priv->mm.stolen, node,
 					  size, alignment, 0,
@@ -343,7 +347,6 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 {
 	resource_size_t reserved_base, stolen_top;
 	resource_size_t reserved_total, reserved_size;
-	resource_size_t stolen_usable_start;
 
 	mutex_init(&dev_priv->mm.stolen_lock);
 
@@ -435,17 +438,11 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 			 (u64)resource_size(&dev_priv->dsm) >> 10,
 			 ((u64)resource_size(&dev_priv->dsm) - reserved_total) >> 10);
 
-	stolen_usable_start = 0;
-	/* WaSkipStolenMemoryFirstPage:bdw+ */
-	if (INTEL_GEN(dev_priv) >= 8)
-		stolen_usable_start = 4096;
-
 	dev_priv->stolen_usable_size =
-		resource_size(&dev_priv->dsm) - reserved_total - stolen_usable_start;
+		resource_size(&dev_priv->dsm) - reserved_total;
 
 	/* Basic memrange allocator for stolen space. */
-	drm_mm_init(&dev_priv->mm.stolen, stolen_usable_start,
-		    dev_priv->stolen_usable_size);
+	drm_mm_init(&dev_priv->mm.stolen, 0, dev_priv->stolen_usable_size);
 
 	return 0;
 }

drivers/gpu/drm/i915/i915_gem_timeline.c

-/*
- * Copyright © 2016 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#include "i915_drv.h"
-#include "i915_syncmap.h"
-
-static void __intel_timeline_init(struct intel_timeline *tl,
-				  struct i915_gem_timeline *parent,
-				  u64 context,
-				  struct lock_class_key *lockclass,
-				  const char *lockname)
-{
-	tl->fence_context = context;
-	tl->common = parent;
-	spin_lock_init(&tl->lock);
-	lockdep_set_class_and_name(&tl->lock, lockclass, lockname);
-	init_request_active(&tl->last_request, NULL);
-	INIT_LIST_HEAD(&tl->requests);
-	i915_syncmap_init(&tl->sync);
-}
-
-static void __intel_timeline_fini(struct intel_timeline *tl)
-{
-	GEM_BUG_ON(!list_empty(&tl->requests));
-
-	i915_syncmap_free(&tl->sync);
-}
-
-static int __i915_gem_timeline_init(struct drm_i915_private *i915,
-				    struct i915_gem_timeline *timeline,
-				    const char *name,
-				    struct lock_class_key *lockclass,
-				    const char *lockname)
-{
-	unsigned int i;
-	u64 fences;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	/*
-	 * Ideally we want a set of engines on a single leaf as we expect
-	 * to mostly be tracking synchronisation between engines. It is not
-	 * a huge issue if this is not the case, but we may want to mitigate
-	 * any page crossing penalties if they become an issue.
-	 */
-	BUILD_BUG_ON(KSYNCMAP < I915_NUM_ENGINES);
-
-	timeline->i915 = i915;
-	timeline->name = kstrdup(name ?: "[kernel]", GFP_KERNEL);
-	if (!timeline->name)
-		return -ENOMEM;
-
-	list_add(&timeline->link, &i915->gt.timelines);
-
-	/* Called during early_init before we know how many engines there are */
-	fences = dma_fence_context_alloc(ARRAY_SIZE(timeline->engine));
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
-		__intel_timeline_init(&timeline->engine[i],
-				      timeline, fences++,
-				      lockclass, lockname);
-
-	return 0;
-}
-
-int i915_gem_timeline_init(struct drm_i915_private *i915,
-			   struct i915_gem_timeline *timeline,
-			   const char *name)
-{
-	static struct lock_class_key class;
-
-	return __i915_gem_timeline_init(i915, timeline, name,
-					&class, "&timeline->lock");
-}
-
-int i915_gem_timeline_init__global(struct drm_i915_private *i915)
-{
-	static struct lock_class_key class;
-
-	return __i915_gem_timeline_init(i915,
-					&i915->gt.global_timeline,
-					"[execution]",
-					&class, "&global_timeline->lock");
-}
-
-/**
- * i915_gem_timelines_park - called when the driver idles
- * @i915: the drm_i915_private device
- *
- * When the driver is completely idle, we know that all of our sync points
- * have been signaled and our tracking is then entirely redundant. Any request
- * to wait upon an older sync point will be completed instantly as we know
- * the fence is signaled and therefore we will not even look them up in the
- * sync point map.
- */
-void i915_gem_timelines_park(struct drm_i915_private *i915)
-{
-	struct i915_gem_timeline *timeline;
-	int i;
-
-	lockdep_assert_held(&i915->drm.struct_mutex);
-
-	list_for_each_entry(timeline, &i915->gt.timelines, link) {
-		for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
-			struct intel_timeline *tl = &timeline->engine[i];
-
-			/*
-			 * All known fences are completed so we can scrap
-			 * the current sync point tracking and start afresh,
-			 * any attempt to wait upon a previous sync point
-			 * will be skipped as the fence was signaled.
-			 */
-			i915_syncmap_free(&tl->sync);
-		}
-	}
-}
-
-void i915_gem_timeline_fini(struct i915_gem_timeline *timeline)
-{
-	int i;
-
-	lockdep_assert_held(&timeline->i915->drm.struct_mutex);
-
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
-		__intel_timeline_fini(&timeline->engine[i]);
-
-	list_del(&timeline->link);
-	kfree(timeline->name);
-}
-
-#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
-#include "selftests/mock_timeline.c"
-#include "selftests/i915_gem_timeline.c"
-#endif

drivers/gpu/drm/i915/i915_gem_userptr.c

@@ -778,6 +778,9 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 			    I915_USERPTR_UNSYNCHRONIZED))
 		return -EINVAL;
 
+	if (!args->user_size)
+		return -EINVAL;
+
 	if (offset_in_page(args->user_ptr | args->user_size))
 		return -EINVAL;
 

drivers/gpu/drm/i915/i915_gpu_error.c

@@ -404,16 +404,17 @@ static const char *bannable(const struct drm_i915_error_context *ctx)
 
 static void error_print_request(struct drm_i915_error_state_buf *m,
 				const char *prefix,
-				const struct drm_i915_error_request *erq)
+				const struct drm_i915_error_request *erq,
+				const unsigned long epoch)
 {
 	if (!erq->seqno)
 		return;
 
-	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms ago, head %08x, tail %08x\n",
+	err_printf(m, "%s pid %d, ban score %d, seqno %8x:%08x, prio %d, emitted %dms, start %08x, head %08x, tail %08x\n",
 		   prefix, erq->pid, erq->ban_score,
-		   erq->context, erq->seqno, erq->priority,
-		   jiffies_to_msecs(jiffies - erq->jiffies),
-		   erq->head, erq->tail);
+		   erq->context, erq->seqno, erq->sched_attr.priority,
+		   jiffies_to_msecs(erq->jiffies - epoch),
+		   erq->start, erq->head, erq->tail);
 }
 
 static void error_print_context(struct drm_i915_error_state_buf *m,
@@ -422,12 +423,13 @@ static void error_print_context(struct drm_i915_error_state_buf *m,
 {
 	err_printf(m, "%s%s[%d] user_handle %d hw_id %d, prio %d, ban score %d%s guilty %d active %d\n",
 		   header, ctx->comm, ctx->pid, ctx->handle, ctx->hw_id,
-		   ctx->priority, ctx->ban_score, bannable(ctx),
+		   ctx->sched_attr.priority, ctx->ban_score, bannable(ctx),
 		   ctx->guilty, ctx->active);
 }
 
 static void error_print_engine(struct drm_i915_error_state_buf *m,
-			       const struct drm_i915_error_engine *ee)
+			       const struct drm_i915_error_engine *ee,
+			       const unsigned long epoch)
 {
 	int n;
 
@@ -497,14 +499,15 @@ static void error_print_engine(struct drm_i915_error_state_buf *m,
 	err_printf(m, "  hangcheck stall: %s\n", yesno(ee->hangcheck_stalled));
 	err_printf(m, "  hangcheck action: %s\n",
 		   hangcheck_action_to_str(ee->hangcheck_action));
-	err_printf(m, "  hangcheck action timestamp: %lu, %u ms ago\n",
+	err_printf(m, "  hangcheck action timestamp: %dms (%lu%s)\n",
+		   jiffies_to_msecs(ee->hangcheck_timestamp - epoch),
 		   ee->hangcheck_timestamp,
-		   jiffies_to_msecs(jiffies - ee->hangcheck_timestamp));
+		   ee->hangcheck_timestamp == epoch ? "; epoch" : "");
 	err_printf(m, "  engine reset count: %u\n", ee->reset_count);
 
 	for (n = 0; n < ee->num_ports; n++) {
 		err_printf(m, "  ELSP[%d]:", n);
-		error_print_request(m, " ", &ee->execlist[n]);
+		error_print_request(m, " ", &ee->execlist[n], epoch);
 	}
 
 	error_print_context(m, "  Active context: ", &ee->context);
@@ -650,6 +653,11 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	ts = ktime_to_timespec64(error->uptime);
 	err_printf(m, "Uptime: %lld s %ld us\n",
 		   (s64)ts.tv_sec, ts.tv_nsec / NSEC_PER_USEC);
+	err_printf(m, "Epoch: %lu jiffies (%u HZ)\n", error->epoch, HZ);
+	err_printf(m, "Capture: %lu jiffies; %d ms ago, %d ms after epoch\n",
+		   error->capture,
+		   jiffies_to_msecs(jiffies - error->capture),
+		   jiffies_to_msecs(error->capture - error->epoch));
 
 	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
 		if (error->engine[i].hangcheck_stalled &&
@@ -710,7 +718,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 
 	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
 		if (error->engine[i].engine_id != -1)
-			error_print_engine(m, &error->engine[i]);
+			error_print_engine(m, &error->engine[i], error->epoch);
 	}
 
 	for (i = 0; i < ARRAY_SIZE(error->active_vm); i++) {
@@ -769,7 +777,9 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 				   dev_priv->engine[i]->name,
 				   ee->num_requests);
 			for (j = 0; j < ee->num_requests; j++)
-				error_print_request(m, " ", &ee->requests[j]);
+				error_print_request(m, " ",
+						    &ee->requests[j],
+						    error->epoch);
 		}
 
 		if (IS_ERR(ee->waiters)) {
@@ -1278,10 +1288,11 @@ static void record_request(struct i915_request *request,
 			   struct drm_i915_error_request *erq)
 {
 	erq->context = request->ctx->hw_id;
-	erq->priority = request->priotree.priority;
+	erq->sched_attr = request->sched.attr;
 	erq->ban_score = atomic_read(&request->ctx->ban_score);
 	erq->seqno = request->global_seqno;
 	erq->jiffies = request->emitted_jiffies;
+	erq->start = i915_ggtt_offset(request->ring->vma);
 	erq->head = request->head;
 	erq->tail = request->tail;
 
@@ -1299,7 +1310,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 
 	count = 0;
 	request = first;
-	list_for_each_entry_from(request, &engine->timeline->requests, link)
+	list_for_each_entry_from(request, &engine->timeline.requests, link)
 		count++;
 	if (!count)
 		return;
@@ -1312,7 +1323,7 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 
 	count = 0;
 	request = first;
-	list_for_each_entry_from(request, &engine->timeline->requests, link) {
+	list_for_each_entry_from(request, &engine->timeline.requests, link) {
 		if (count >= ee->num_requests) {
 			/*
 			 * If the ring request list was changed in
@@ -1372,7 +1383,7 @@ static void record_context(struct drm_i915_error_context *e,
 
 	e->handle = ctx->user_handle;
 	e->hw_id = ctx->hw_id;
-	e->priority = ctx->priority;
+	e->sched_attr = ctx->sched;
 	e->ban_score = atomic_read(&ctx->ban_score);
 	e->bannable = i915_gem_context_is_bannable(ctx);
 	e->guilty = atomic_read(&ctx->guilty_count);
@@ -1472,7 +1483,8 @@ static void gem_record_rings(struct i915_gpu_state *error)
 
 			ee->ctx =
 				i915_error_object_create(i915,
-							 request->ctx->engine[i].state);
+							 to_intel_context(request->ctx,
+									  engine)->state);
 
 			error->simulated |=
 				i915_gem_context_no_error_capture(request->ctx);
@@ -1735,6 +1747,22 @@ static void capture_params(struct i915_gpu_state *error)
 #undef DUP
 }
 
+static unsigned long capture_find_epoch(const struct i915_gpu_state *error)
+{
+	unsigned long epoch = error->capture;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		const struct drm_i915_error_engine *ee = &error->engine[i];
+
+		if (ee->hangcheck_stalled &&
+		    time_before(ee->hangcheck_timestamp, epoch))
+			epoch = ee->hangcheck_timestamp;
+	}
+
+	return epoch;
+}
+
 static int capture(void *data)
 {
 	struct i915_gpu_state *error = data;
@@ -1743,6 +1771,7 @@ static int capture(void *data)
 	error->boottime = ktime_get_boottime();
 	error->uptime = ktime_sub(ktime_get(),
 				  error->i915->gt.last_init_time);
+	error->capture = jiffies;
 
 	capture_params(error);
 	capture_gen_state(error);
@@ -1756,6 +1785,8 @@ static int capture(void *data)
 	error->overlay = intel_overlay_capture_error_state(error->i915);
 	error->display = intel_display_capture_error_state(error->i915);
 
+	error->epoch = capture_find_epoch(error);
+
 	return 0;
 }
 

drivers/gpu/drm/i915/i915_gpu_error.h

@@ -20,6 +20,7 @@
 #include "i915_gem.h"
 #include "i915_gem_gtt.h"
 #include "i915_params.h"
+#include "i915_scheduler.h"
 
 struct drm_i915_private;
 struct intel_overlay_error_state;
@@ -30,6 +31,8 @@ struct i915_gpu_state {
 	ktime_t time;
 	ktime_t boottime;
 	ktime_t uptime;
+	unsigned long capture;
+	unsigned long epoch;
 
 	struct drm_i915_private *i915;
 
@@ -122,11 +125,11 @@ struct i915_gpu_state {
 			pid_t pid;
 			u32 handle;
 			u32 hw_id;
-			int priority;
 			int ban_score;
 			int active;
 			int guilty;
 			bool bannable;
+			struct i915_sched_attr sched_attr;
 		} context;
 
 		struct drm_i915_error_object {
@@ -147,11 +150,12 @@ struct i915_gpu_state {
 			long jiffies;
 			pid_t pid;
 			u32 context;
-			int priority;
 			int ban_score;
 			u32 seqno;
+			u32 start;
 			u32 head;
 			u32 tail;
+			struct i915_sched_attr sched_attr;
 		} *requests, execlist[EXECLIST_MAX_PORTS];
 		unsigned int num_ports;
 

drivers/gpu/drm/i915/i915_irq.c

@@ -247,9 +247,9 @@ static u32
 gen11_gt_engine_identity(struct drm_i915_private * const i915,
 			 const unsigned int bank, const unsigned int bit);
 
-static bool gen11_reset_one_iir(struct drm_i915_private * const i915,
-				const unsigned int bank,
-				const unsigned int bit)
+bool gen11_reset_one_iir(struct drm_i915_private * const i915,
+			 const unsigned int bank,
+			 const unsigned int bit)
 {
 	void __iomem * const regs = i915->regs;
 	u32 dw;
@@ -2464,6 +2464,13 @@ static void ivb_display_irq_handler(struct drm_i915_private *dev_priv,
 	if (de_iir & DE_ERR_INT_IVB)
 		ivb_err_int_handler(dev_priv);
 
+	if (de_iir & DE_EDP_PSR_INT_HSW) {
+		u32 psr_iir = I915_READ(EDP_PSR_IIR);
+
+		intel_psr_irq_handler(dev_priv, psr_iir);
+		I915_WRITE(EDP_PSR_IIR, psr_iir);
+	}
+
 	if (de_iir & DE_AUX_CHANNEL_A_IVB)
 		dp_aux_irq_handler(dev_priv);
 
@@ -2593,11 +2600,25 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 	if (master_ctl & GEN8_DE_MISC_IRQ) {
 		iir = I915_READ(GEN8_DE_MISC_IIR);
 		if (iir) {
+			bool found = false;
+
 			I915_WRITE(GEN8_DE_MISC_IIR, iir);
 			ret = IRQ_HANDLED;
-			if (iir & GEN8_DE_MISC_GSE)
+
+			if (iir & GEN8_DE_MISC_GSE) {
 				intel_opregion_asle_intr(dev_priv);
-			else
+				found = true;
+			}
+
+			if (iir & GEN8_DE_EDP_PSR) {
+				u32 psr_iir = I915_READ(EDP_PSR_IIR);
+
+				intel_psr_irq_handler(dev_priv, psr_iir);
+				I915_WRITE(EDP_PSR_IIR, psr_iir);
+				found = true;
+			}
+
+			if (!found)
 				DRM_ERROR("Unexpected DE Misc interrupt\n");
 		}
 		else
@@ -3348,6 +3369,11 @@ static void ironlake_irq_reset(struct drm_device *dev)
 	if (IS_GEN7(dev_priv))
 		I915_WRITE(GEN7_ERR_INT, 0xffffffff);
 
+	if (IS_HASWELL(dev_priv)) {
+		I915_WRITE(EDP_PSR_IMR, 0xffffffff);
+		I915_WRITE(EDP_PSR_IIR, 0xffffffff);
+	}
+
 	gen5_gt_irq_reset(dev_priv);
 
 	ibx_irq_reset(dev_priv);
@@ -3386,6 +3412,9 @@ static void gen8_irq_reset(struct drm_device *dev)
 
 	gen8_gt_irq_reset(dev_priv);
 
+	I915_WRITE(EDP_PSR_IMR, 0xffffffff);
+	I915_WRITE(EDP_PSR_IIR, 0xffffffff);
+
 	for_each_pipe(dev_priv, pipe)
 		if (intel_display_power_is_enabled(dev_priv,
 						   POWER_DOMAIN_PIPE(pipe)))
@@ -3762,6 +3791,12 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 			      DE_DP_A_HOTPLUG);
 	}
 
+	if (IS_HASWELL(dev_priv)) {
+		gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
+		intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
+		display_mask |= DE_EDP_PSR_INT_HSW;
+	}
+
 	dev_priv->irq_mask = ~display_mask;
 
 	ibx_irq_pre_postinstall(dev);
@@ -3872,7 +3907,7 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	uint32_t de_pipe_enables;
 	u32 de_port_masked = GEN8_AUX_CHANNEL_A;
 	u32 de_port_enables;
-	u32 de_misc_masked = GEN8_DE_MISC_GSE;
+	u32 de_misc_masked = GEN8_DE_MISC_GSE | GEN8_DE_EDP_PSR;
 	enum pipe pipe;
 
 	if (INTEL_GEN(dev_priv) >= 9) {
@@ -3897,6 +3932,9 @@ static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 	else if (IS_BROADWELL(dev_priv))
 		de_port_enables |= GEN8_PORT_DP_A_HOTPLUG;
 
+	gen3_assert_iir_is_zero(dev_priv, EDP_PSR_IIR);
+	intel_psr_irq_control(dev_priv, dev_priv->psr.debug);
+
 	for_each_pipe(dev_priv, pipe) {
 		dev_priv->de_irq_mask[pipe] = ~de_pipe_masked;
 

drivers/gpu/drm/i915/i915_params.c

@@ -164,6 +164,9 @@ i915_param_named_unsafe(guc_firmware_path, charp, 0400,
 i915_param_named_unsafe(huc_firmware_path, charp, 0400,
 	"HuC firmware path to use instead of the default one");
 
+i915_param_named_unsafe(dmc_firmware_path, charp, 0400,
+	"DMC firmware path to use instead of the default one");
+
 i915_param_named_unsafe(enable_dp_mst, bool, 0600,
 	"Enable multi-stream transport (MST) for new DisplayPort sinks. (default: true)");
 

drivers/gpu/drm/i915/i915_params.h

@@ -51,6 +51,7 @@ struct drm_printer;
 	param(int, guc_log_level, -1) \
 	param(char *, guc_firmware_path, NULL) \
 	param(char *, huc_firmware_path, NULL) \
+	param(char *, dmc_firmware_path, NULL) \
 	param(int, mmio_debug, 0) \
 	param(int, edp_vswing, 0) \
 	param(int, reset, 2) \

drivers/gpu/drm/i915/i915_perf.c

@@ -1234,7 +1234,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 		 *
 		 * NB: implied RCS engine...
 		 */
-		ring = engine->context_pin(engine, stream->ctx);
+		ring = intel_context_pin(stream->ctx, engine);
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 		if (IS_ERR(ring))
 			return PTR_ERR(ring);
@@ -1246,7 +1246,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 		 * with gen8+ and execlists
 		 */
 		dev_priv->perf.oa.specific_ctx_id =
-			i915_ggtt_offset(stream->ctx->engine[engine->id].state);
+			i915_ggtt_offset(to_intel_context(stream->ctx, engine)->state);
 	}
 
 	return 0;
@@ -1271,7 +1271,7 @@ static void oa_put_render_ctx_id(struct i915_perf_stream *stream)
 		mutex_lock(&dev_priv->drm.struct_mutex);
 
 		dev_priv->perf.oa.specific_ctx_id = INVALID_CTX_ID;
-		engine->context_unpin(engine, stream->ctx);
+		intel_context_unpin(stream->ctx, engine);
 
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
@@ -1695,7 +1695,7 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 						 const struct i915_oa_config *oa_config)
 {
 	struct intel_engine_cs *engine = dev_priv->engine[RCS];
-	struct i915_gem_timeline *timeline;
+	struct i915_timeline *timeline;
 	struct i915_request *rq;
 	int ret;
 
@@ -1716,15 +1716,11 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 	/* Queue this switch after all other activity */
 	list_for_each_entry(timeline, &dev_priv->gt.timelines, link) {
 		struct i915_request *prev;
-		struct intel_timeline *tl;
 
-		tl = &timeline->engine[engine->id];
-		prev = i915_gem_active_raw(&tl->last_request,
+		prev = i915_gem_active_raw(&timeline->last_request,
 					   &dev_priv->drm.struct_mutex);
 		if (prev)
-			i915_sw_fence_await_sw_fence_gfp(&rq->submit,
-							 &prev->submit,
-							 GFP_KERNEL);
+			i915_request_await_dma_fence(rq, &prev->fence);
 	}
 
 	i915_request_add(rq);
@@ -1759,6 +1755,7 @@ static int gen8_switch_to_updated_kernel_context(struct drm_i915_private *dev_pr
 static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv,
 				       const struct i915_oa_config *oa_config)
 {
+	struct intel_engine_cs *engine = dev_priv->engine[RCS];
 	struct i915_gem_context *ctx;
 	int ret;
 	unsigned int wait_flags = I915_WAIT_LOCKED;
@@ -1789,7 +1786,7 @@ static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv,
 
 	/* Update all contexts now that we've stalled the submission. */
 	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
-		struct intel_context *ce = &ctx->engine[RCS];
+		struct intel_context *ce = to_intel_context(ctx, engine);
 		u32 *regs;
 
 		/* OA settings will be set upon first use */
@@ -1963,11 +1960,19 @@ static void i915_oa_stream_enable(struct i915_perf_stream *stream)
 static void gen7_oa_disable(struct drm_i915_private *dev_priv)
 {
 	I915_WRITE(GEN7_OACONTROL, 0);
+	if (intel_wait_for_register(dev_priv,
+				    GEN7_OACONTROL, GEN7_OACONTROL_ENABLE, 0,
+				    50))
+		DRM_ERROR("wait for OA to be disabled timed out\n");
 }
 
 static void gen8_oa_disable(struct drm_i915_private *dev_priv)
 {
 	I915_WRITE(GEN8_OACONTROL, 0);
+	if (intel_wait_for_register(dev_priv,
+				    GEN8_OACONTROL, GEN8_OA_COUNTER_ENABLE, 0,
+				    50))
+		DRM_ERROR("wait for OA to be disabled timed out\n");
 }
 
 /**

drivers/gpu/drm/i915/i915_reg.h

@@ -3840,6 +3840,7 @@ enum {
 #define SLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x94d4)
 #define  SARBUNIT_CLKGATE_DIS		(1 << 5)
 #define  RCCUNIT_CLKGATE_DIS		(1 << 7)
+#define  MSCUNIT_CLKGATE_DIS		(1 << 10)
 
 #define SUBSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9524)
 #define  GWUNIT_CLKGATE_DIS		(1 << 16)
@@ -3847,6 +3848,9 @@ enum {
 #define UNSLICE_UNIT_LEVEL_CLKGATE	_MMIO(0x9434)
 #define  VFUNIT_CLKGATE_DIS		(1 << 20)
 
+#define INF_UNIT_LEVEL_CLKGATE		_MMIO(0x9560)
+#define   CGPSF_CLKGATE_DIS		(1 << 3)
+
 /*
  * Display engine regs
  */
@@ -4026,6 +4030,13 @@ enum {
 #define   EDP_PSR_TP1_TIME_0us			(3<<4)
 #define   EDP_PSR_IDLE_FRAME_SHIFT		0
 
+/* Bspec claims those aren't shifted but stay at 0x64800 */
+#define EDP_PSR_IMR				_MMIO(0x64834)
+#define EDP_PSR_IIR				_MMIO(0x64838)
+#define   EDP_PSR_ERROR(trans)			(1 << (((trans) * 8 + 10) & 31))
+#define   EDP_PSR_POST_EXIT(trans)		(1 << (((trans) * 8 + 9) & 31))
+#define   EDP_PSR_PRE_ENTRY(trans)		(1 << (((trans) * 8 + 8) & 31))
+
 #define EDP_PSR_AUX_CTL				_MMIO(dev_priv->psr_mmio_base + 0x10)
 #define   EDP_PSR_AUX_CTL_TIME_OUT_MASK		(3 << 26)
 #define   EDP_PSR_AUX_CTL_MESSAGE_SIZE_MASK	(0x1f << 20)
@@ -4088,6 +4099,29 @@ enum {
 #define   EDP_PSR2_IDLE_FRAME_MASK	0xf
 #define   EDP_PSR2_IDLE_FRAME_SHIFT	0
 
+#define _PSR_EVENT_TRANS_A			0x60848
+#define _PSR_EVENT_TRANS_B			0x61848
+#define _PSR_EVENT_TRANS_C			0x62848
+#define _PSR_EVENT_TRANS_D			0x63848
+#define _PSR_EVENT_TRANS_EDP			0x6F848
+#define PSR_EVENT(trans)			_MMIO_TRANS2(trans, _PSR_EVENT_TRANS_A)
+#define  PSR_EVENT_PSR2_WD_TIMER_EXPIRE		(1 << 17)
+#define  PSR_EVENT_PSR2_DISABLED		(1 << 16)
+#define  PSR_EVENT_SU_DIRTY_FIFO_UNDERRUN	(1 << 15)
+#define  PSR_EVENT_SU_CRC_FIFO_UNDERRUN		(1 << 14)
+#define  PSR_EVENT_GRAPHICS_RESET		(1 << 12)
+#define  PSR_EVENT_PCH_INTERRUPT		(1 << 11)
+#define  PSR_EVENT_MEMORY_UP			(1 << 10)
+#define  PSR_EVENT_FRONT_BUFFER_MODIFY		(1 << 9)
+#define  PSR_EVENT_WD_TIMER_EXPIRE		(1 << 8)
+#define  PSR_EVENT_PIPE_REGISTERS_UPDATE	(1 << 6)
+#define  PSR_EVENT_REGISTER_UPDATE		(1 << 5)
+#define  PSR_EVENT_HDCP_ENABLE			(1 << 4)
+#define  PSR_EVENT_KVMR_SESSION_ENABLE		(1 << 3)
+#define  PSR_EVENT_VBI_ENABLE			(1 << 2)
+#define  PSR_EVENT_LPSP_MODE_EXIT		(1 << 1)
+#define  PSR_EVENT_PSR_DISABLE			(1 << 0)
+
 #define EDP_PSR2_STATUS			_MMIO(0x6f940)
 #define EDP_PSR2_STATUS_STATE_MASK     (0xf<<28)
 #define EDP_PSR2_STATUS_STATE_SHIFT    28
@@ -6377,9 +6411,9 @@ enum {
 #define _PLANE_COLOR_CTL_1_A			0x701CC /* GLK+ */
 #define _PLANE_COLOR_CTL_2_A			0x702CC /* GLK+ */
 #define _PLANE_COLOR_CTL_3_A			0x703CC /* GLK+ */
-#define   PLANE_COLOR_PIPE_GAMMA_ENABLE		(1 << 30)
+#define   PLANE_COLOR_PIPE_GAMMA_ENABLE		(1 << 30) /* Pre-ICL */
 #define   PLANE_COLOR_YUV_RANGE_CORRECTION_DISABLE	(1 << 28)
-#define   PLANE_COLOR_PIPE_CSC_ENABLE		(1 << 23)
+#define   PLANE_COLOR_PIPE_CSC_ENABLE		(1 << 23) /* Pre-ICL */
 #define   PLANE_COLOR_CSC_MODE_BYPASS			(0 << 17)
 #define   PLANE_COLOR_CSC_MODE_YUV601_TO_RGB709		(1 << 17)
 #define   PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709		(2 << 17)
@@ -6474,6 +6508,9 @@ enum {
 
 #define _PLANE_BUF_CFG_1_B			0x7127c
 #define _PLANE_BUF_CFG_2_B			0x7137c
+#define  SKL_DDB_ENTRY_MASK			0x3FF
+#define  ICL_DDB_ENTRY_MASK			0x7FF
+#define  DDB_ENTRY_END_SHIFT			16
 #define _PLANE_BUF_CFG_1(pipe)	\
 	_PIPE(pipe, _PLANE_BUF_CFG_1_A, _PLANE_BUF_CFG_1_B)
 #define _PLANE_BUF_CFG_2(pipe)	\
@@ -6837,6 +6874,7 @@ enum {
 #define DE_PCH_EVENT_IVB		(1<<28)
 #define DE_DP_A_HOTPLUG_IVB		(1<<27)
 #define DE_AUX_CHANNEL_A_IVB		(1<<26)
+#define DE_EDP_PSR_INT_HSW		(1<<19)
 #define DE_SPRITEC_FLIP_DONE_IVB	(1<<14)
 #define DE_PLANEC_FLIP_DONE_IVB		(1<<13)
 #define DE_PIPEC_VBLANK_IVB		(1<<10)
@@ -6961,6 +6999,7 @@ enum {
 #define GEN8_DE_MISC_IIR _MMIO(0x44468)
 #define GEN8_DE_MISC_IER _MMIO(0x4446c)
 #define  GEN8_DE_MISC_GSE		(1 << 27)
+#define  GEN8_DE_EDP_PSR		(1 << 19)
 
 #define GEN8_PCU_ISR _MMIO(0x444e0)
 #define GEN8_PCU_IMR _MMIO(0x444e4)
@@ -7191,18 +7230,22 @@ enum {
 #define GEN7_L3CNTLREG3				_MMIO(0xB024)
 
 #define GEN7_L3_CHICKEN_MODE_REGISTER		_MMIO(0xB030)
-#define  GEN7_WA_L3_CHICKEN_MODE				0x20000000
+#define   GEN7_WA_L3_CHICKEN_MODE		0x20000000
+#define GEN10_L3_CHICKEN_MODE_REGISTER		_MMIO(0xB114)
+#define   GEN11_I2M_WRITE_DISABLE		(1 << 28)
 
 #define GEN7_L3SQCREG4				_MMIO(0xb034)
 #define  L3SQ_URB_READ_CAM_MATCH_DISABLE	(1<<27)
 
 #define GEN8_L3SQCREG4				_MMIO(0xb118)
-#define  GEN8_LQSC_RO_PERF_DIS			(1<<27)
-#define  GEN8_LQSC_FLUSH_COHERENT_LINES		(1<<21)
+#define  GEN11_LQSC_CLEAN_EVICT_DISABLE		(1 << 6)
+#define  GEN8_LQSC_RO_PERF_DIS			(1 << 27)
+#define  GEN8_LQSC_FLUSH_COHERENT_LINES		(1 << 21)
 
 /* GEN8 chicken */
 #define HDC_CHICKEN0				_MMIO(0x7300)
 #define CNL_HDC_CHICKEN0			_MMIO(0xE5F0)
+#define ICL_HDC_MODE				_MMIO(0xE5F4)
 #define  HDC_FORCE_CSR_NON_COHERENT_OVR_DISABLE	(1<<15)
 #define  HDC_FENCE_DEST_SLM_DISABLE		(1<<14)
 #define  HDC_DONOT_FETCH_MEM_WHEN_MASKED	(1<<11)
@@ -7216,6 +7259,9 @@ enum {
 #define SLICE_ECO_CHICKEN0			_MMIO(0x7308)
 #define   PIXEL_MASK_CAMMING_DISABLE		(1 << 14)
 
+#define GEN9_WM_CHICKEN3			_MMIO(0x5588)
+#define   GEN9_FACTOR_IN_CLR_VAL_HIZ		(1 << 9)
+
 /* WaCatErrorRejectionIssue */
 #define GEN7_SQ_CHICKEN_MBCUNIT_CONFIG		_MMIO(0x9030)
 #define  GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB	(1<<11)
@@ -8214,8 +8260,30 @@ enum {
 #define   GEN8_DOP_CLOCK_GATE_GUC_ENABLE	(1<<4)
 #define   GEN8_DOP_CLOCK_GATE_MEDIA_ENABLE     (1<<6)
 
-#define GEN8_GARBCNTL                   _MMIO(0xB004)
-#define   GEN9_GAPS_TSV_CREDIT_DISABLE  (1<<7)
+#define GEN8_GARBCNTL				_MMIO(0xB004)
+#define   GEN9_GAPS_TSV_CREDIT_DISABLE		(1 << 7)
+#define   GEN11_ARBITRATION_PRIO_ORDER_MASK	(0x3f << 22)
+#define   GEN11_HASH_CTRL_EXCL_MASK		(0x7f << 0)
+#define   GEN11_HASH_CTRL_EXCL_BIT0		(1 << 0)
+
+#define GEN11_GLBLINVL				_MMIO(0xB404)
+#define   GEN11_BANK_HASH_ADDR_EXCL_MASK	(0x7f << 5)
+#define   GEN11_BANK_HASH_ADDR_EXCL_BIT0	(1 << 5)
+
+#define GEN10_DFR_RATIO_EN_AND_CHICKEN	_MMIO(0x9550)
+#define   DFR_DISABLE			(1 << 9)
+
+#define GEN11_GACB_PERF_CTRL			_MMIO(0x4B80)
+#define   GEN11_HASH_CTRL_MASK			(0x3 << 12 | 0xf << 0)
+#define   GEN11_HASH_CTRL_BIT0			(1 << 0)
+#define   GEN11_HASH_CTRL_BIT4			(1 << 12)
+
+#define GEN11_LSN_UNSLCVC				_MMIO(0xB43C)
+#define   GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC	(1 << 9)
+#define   GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC	(1 << 7)
+
+#define GAMW_ECO_DEV_RW_IA_REG			_MMIO(0x4080)
+#define   GAMW_ECO_DEV_CTX_RELOAD_DISABLE	(1 << 7)
 
 /* IVYBRIDGE DPF */
 #define GEN7_L3CDERRST1(slice)		_MMIO(0xB008 + (slice) * 0x200) /* L3CD Error Status 1 */
@@ -8724,6 +8792,12 @@ enum skl_power_gate {
 #define  PORT_CLK_SEL_NONE		(7<<29)
 #define  PORT_CLK_SEL_MASK		(7<<29)
 
+/* On ICL+ this is the same as PORT_CLK_SEL, but all bits change. */
+#define DDI_CLK_SEL(port)		PORT_CLK_SEL(port)
+#define  DDI_CLK_SEL_NONE		(0x0 << 28)
+#define  DDI_CLK_SEL_MG			(0x8 << 28)
+#define  DDI_CLK_SEL_MASK		(0xF << 28)
+
 /* Transcoder clock selection */
 #define _TRANS_CLK_SEL_A		0x46140
 #define _TRANS_CLK_SEL_B		0x46144
@@ -8854,6 +8928,7 @@ enum skl_power_gate {
  * CNL Clocks
  */
 #define DPCLKA_CFGCR0				_MMIO(0x6C200)
+#define DPCLKA_CFGCR0_ICL			_MMIO(0x164280)
 #define  DPCLKA_CFGCR0_DDI_CLK_OFF(port)	(1 << ((port) ==  PORT_F ? 23 : \
 						      (port)+10))
 #define  DPCLKA_CFGCR0_DDI_CLK_SEL_SHIFT(port)	((port) == PORT_F ? 21 : \
@@ -8870,10 +8945,141 @@ enum skl_power_gate {
 #define  PLL_POWER_STATE	(1 << 26)
 #define CNL_DPLL_ENABLE(pll)	_MMIO_PLL(pll, DPLL0_ENABLE, DPLL1_ENABLE)
 
+#define _MG_PLL1_ENABLE		0x46030
+#define _MG_PLL2_ENABLE		0x46034
+#define _MG_PLL3_ENABLE		0x46038
+#define _MG_PLL4_ENABLE		0x4603C
+/* Bits are the same as DPLL0_ENABLE */
+#define MG_PLL_ENABLE(port)	_MMIO_PORT((port) - PORT_C, _MG_PLL1_ENABLE, \
+					   _MG_PLL2_ENABLE)
+
+#define _MG_REFCLKIN_CTL_PORT1				0x16892C
+#define _MG_REFCLKIN_CTL_PORT2				0x16992C
+#define _MG_REFCLKIN_CTL_PORT3				0x16A92C
+#define _MG_REFCLKIN_CTL_PORT4				0x16B92C
+#define   MG_REFCLKIN_CTL_OD_2_MUX(x)			((x) << 8)
+#define MG_REFCLKIN_CTL(port) _MMIO_PORT((port) - PORT_C, \
+					 _MG_REFCLKIN_CTL_PORT1, \
+					 _MG_REFCLKIN_CTL_PORT2)
+
+#define _MG_CLKTOP2_CORECLKCTL1_PORT1			0x1688D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT2			0x1698D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT3			0x16A8D8
+#define _MG_CLKTOP2_CORECLKCTL1_PORT4			0x16B8D8
+#define   MG_CLKTOP2_CORECLKCTL1_B_DIVRATIO(x)		((x) << 16)
+#define   MG_CLKTOP2_CORECLKCTL1_A_DIVRATIO(x)		((x) << 8)
+#define MG_CLKTOP2_CORECLKCTL1(port) _MMIO_PORT((port) - PORT_C, \
+						_MG_CLKTOP2_CORECLKCTL1_PORT1, \
+						_MG_CLKTOP2_CORECLKCTL1_PORT2)
+
+#define _MG_CLKTOP2_HSCLKCTL_PORT1			0x1688D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT2			0x1698D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT3			0x16A8D4
+#define _MG_CLKTOP2_HSCLKCTL_PORT4			0x16B8D4
+#define   MG_CLKTOP2_HSCLKCTL_CORE_INPUTSEL(x)		((x) << 16)
+#define   MG_CLKTOP2_HSCLKCTL_TLINEDRV_CLKSEL(x)	((x) << 14)
+#define   MG_CLKTOP2_HSCLKCTL_HSDIV_RATIO(x)		((x) << 12)
+#define   MG_CLKTOP2_HSCLKCTL_DSDIV_RATIO(x)		((x) << 8)
+#define MG_CLKTOP2_HSCLKCTL(port) _MMIO_PORT((port) - PORT_C, \
+					     _MG_CLKTOP2_HSCLKCTL_PORT1, \
+					     _MG_CLKTOP2_HSCLKCTL_PORT2)
+
+#define _MG_PLL_DIV0_PORT1				0x168A00
+#define _MG_PLL_DIV0_PORT2				0x169A00
+#define _MG_PLL_DIV0_PORT3				0x16AA00
+#define _MG_PLL_DIV0_PORT4				0x16BA00
+#define   MG_PLL_DIV0_FRACNEN_H				(1 << 30)
+#define   MG_PLL_DIV0_FBDIV_FRAC(x)			((x) << 8)
+#define   MG_PLL_DIV0_FBDIV_INT(x)			((x) << 0)
+#define MG_PLL_DIV0(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_DIV0_PORT1, \
+				     _MG_PLL_DIV0_PORT2)
+
+#define _MG_PLL_DIV1_PORT1				0x168A04
+#define _MG_PLL_DIV1_PORT2				0x169A04
+#define _MG_PLL_DIV1_PORT3				0x16AA04
+#define _MG_PLL_DIV1_PORT4				0x16BA04
+#define   MG_PLL_DIV1_IREF_NDIVRATIO(x)			((x) << 16)
+#define   MG_PLL_DIV1_DITHER_DIV_1			(0 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_2			(1 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_4			(2 << 12)
+#define   MG_PLL_DIV1_DITHER_DIV_8			(3 << 12)
+#define   MG_PLL_DIV1_NDIVRATIO(x)			((x) << 4)
+#define   MG_PLL_DIV1_FBPREDIV(x)			((x) << 0)
+#define MG_PLL_DIV1(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_DIV1_PORT1, \
+				     _MG_PLL_DIV1_PORT2)
+
+#define _MG_PLL_LF_PORT1				0x168A08
+#define _MG_PLL_LF_PORT2				0x169A08
+#define _MG_PLL_LF_PORT3				0x16AA08
+#define _MG_PLL_LF_PORT4				0x16BA08
+#define   MG_PLL_LF_TDCTARGETCNT(x)			((x) << 24)
+#define   MG_PLL_LF_AFCCNTSEL_256			(0 << 20)
+#define   MG_PLL_LF_AFCCNTSEL_512			(1 << 20)
+#define   MG_PLL_LF_GAINCTRL(x)				((x) << 16)
+#define   MG_PLL_LF_INT_COEFF(x)			((x) << 8)
+#define   MG_PLL_LF_PROP_COEFF(x)			((x) << 0)
+#define MG_PLL_LF(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_LF_PORT1, \
+				   _MG_PLL_LF_PORT2)
+
+#define _MG_PLL_FRAC_LOCK_PORT1				0x168A0C
+#define _MG_PLL_FRAC_LOCK_PORT2				0x169A0C
+#define _MG_PLL_FRAC_LOCK_PORT3				0x16AA0C
+#define _MG_PLL_FRAC_LOCK_PORT4				0x16BA0C
+#define   MG_PLL_FRAC_LOCK_TRUELOCK_CRIT_32		(1 << 18)
+#define   MG_PLL_FRAC_LOCK_EARLYLOCK_CRIT_32		(1 << 16)
+#define   MG_PLL_FRAC_LOCK_LOCKTHRESH(x)		((x) << 11)
+#define   MG_PLL_FRAC_LOCK_DCODITHEREN			(1 << 10)
+#define   MG_PLL_FRAC_LOCK_FEEDFWRDCAL_EN		(1 << 8)
+#define   MG_PLL_FRAC_LOCK_FEEDFWRDGAIN(x)		((x) << 0)
+#define MG_PLL_FRAC_LOCK(port) _MMIO_PORT((port) - PORT_C, \
+					  _MG_PLL_FRAC_LOCK_PORT1, \
+					  _MG_PLL_FRAC_LOCK_PORT2)
+
+#define _MG_PLL_SSC_PORT1				0x168A10
+#define _MG_PLL_SSC_PORT2				0x169A10
+#define _MG_PLL_SSC_PORT3				0x16AA10
+#define _MG_PLL_SSC_PORT4				0x16BA10
+#define   MG_PLL_SSC_EN					(1 << 28)
+#define   MG_PLL_SSC_TYPE(x)				((x) << 26)
+#define   MG_PLL_SSC_STEPLENGTH(x)			((x) << 16)
+#define   MG_PLL_SSC_STEPNUM(x)				((x) << 10)
+#define   MG_PLL_SSC_FLLEN				(1 << 9)
+#define   MG_PLL_SSC_STEPSIZE(x)			((x) << 0)
+#define MG_PLL_SSC(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_SSC_PORT1, \
+				    _MG_PLL_SSC_PORT2)
+
+#define _MG_PLL_BIAS_PORT1				0x168A14
+#define _MG_PLL_BIAS_PORT2				0x169A14
+#define _MG_PLL_BIAS_PORT3				0x16AA14
+#define _MG_PLL_BIAS_PORT4				0x16BA14
+#define   MG_PLL_BIAS_BIAS_GB_SEL(x)			((x) << 30)
+#define   MG_PLL_BIAS_INIT_DCOAMP(x)			((x) << 24)
+#define   MG_PLL_BIAS_BIAS_BONUS(x)			((x) << 16)
+#define   MG_PLL_BIAS_BIASCAL_EN			(1 << 15)
+#define   MG_PLL_BIAS_CTRIM(x)				((x) << 8)
+#define   MG_PLL_BIAS_VREF_RDAC(x)			((x) << 5)
+#define   MG_PLL_BIAS_IREFTRIM(x)			((x) << 0)
+#define MG_PLL_BIAS(port) _MMIO_PORT((port) - PORT_C, _MG_PLL_BIAS_PORT1, \
+				     _MG_PLL_BIAS_PORT2)
+
+#define _MG_PLL_TDC_COLDST_BIAS_PORT1			0x168A18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT2			0x169A18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT3			0x16AA18
+#define _MG_PLL_TDC_COLDST_BIAS_PORT4			0x16BA18
+#define   MG_PLL_TDC_COLDST_IREFINT_EN			(1 << 27)
+#define   MG_PLL_TDC_COLDST_REFBIAS_START_PULSE_W(x)	((x) << 17)
+#define   MG_PLL_TDC_COLDST_COLDSTART			(1 << 16)
+#define   MG_PLL_TDC_TDCOVCCORR_EN			(1 << 2)
+#define   MG_PLL_TDC_TDCSEL(x)				((x) << 0)
+#define MG_PLL_TDC_COLDST_BIAS(port) _MMIO_PORT((port) - PORT_C, \
+						_MG_PLL_TDC_COLDST_BIAS_PORT1, \
+						_MG_PLL_TDC_COLDST_BIAS_PORT2)
+
 #define _CNL_DPLL0_CFGCR0		0x6C000
 #define _CNL_DPLL1_CFGCR0		0x6C080
 #define  DPLL_CFGCR0_HDMI_MODE		(1 << 30)
 #define  DPLL_CFGCR0_SSC_ENABLE		(1 << 29)
+#define  DPLL_CFGCR0_SSC_ENABLE_ICL	(1 << 25)
 #define  DPLL_CFGCR0_LINK_RATE_MASK	(0xf << 25)
 #define  DPLL_CFGCR0_LINK_RATE_2700	(0 << 25)
 #define  DPLL_CFGCR0_LINK_RATE_1350	(1 << 25)
@@ -8907,8 +9113,19 @@ enum skl_power_gate {
 #define  DPLL_CFGCR1_PDIV_5		(4 << 2)
 #define  DPLL_CFGCR1_PDIV_7		(8 << 2)
 #define  DPLL_CFGCR1_CENTRAL_FREQ	(3 << 0)
+#define  DPLL_CFGCR1_CENTRAL_FREQ_8400	(3 << 0)
 #define CNL_DPLL_CFGCR1(pll)		_MMIO_PLL(pll, _CNL_DPLL0_CFGCR1, _CNL_DPLL1_CFGCR1)
 
+#define _ICL_DPLL0_CFGCR0		0x164000
+#define _ICL_DPLL1_CFGCR0		0x164080
+#define ICL_DPLL_CFGCR0(pll)		_MMIO_PLL(pll, _ICL_DPLL0_CFGCR0, \
+						  _ICL_DPLL1_CFGCR0)
+
+#define _ICL_DPLL0_CFGCR1		0x164004
+#define _ICL_DPLL1_CFGCR1		0x164084
+#define ICL_DPLL_CFGCR1(pll)		_MMIO_PLL(pll, _ICL_DPLL0_CFGCR1, \
+						  _ICL_DPLL1_CFGCR1)
+
 /* BXT display engine PLL */
 #define BXT_DE_PLL_CTL			_MMIO(0x6d000)
 #define   BXT_DE_PLL_RATIO(x)		(x)	/* {60,65,100} * 19.2MHz */
@@ -9680,6 +9897,13 @@ enum skl_power_gate {
 #define GEN9_MFX1_MOCS(i)	_MMIO(0xca00 + (i) * 4)	/* Media 1 MOCS registers */
 #define GEN9_VEBOX_MOCS(i)	_MMIO(0xcb00 + (i) * 4)	/* Video MOCS registers */
 #define GEN9_BLT_MOCS(i)	_MMIO(0xcc00 + (i) * 4)	/* Blitter MOCS registers */
+/* Media decoder 2 MOCS registers */
+#define GEN11_MFX2_MOCS(i)	_MMIO(0x10000 + (i) * 4)
+
+#define GEN10_SCRATCH_LNCF2		_MMIO(0xb0a0)
+#define   PMFLUSHDONE_LNICRSDROP	(1 << 20)
+#define   PMFLUSH_GAPL3UNBLOCK		(1 << 21)
+#define   PMFLUSHDONE_LNEBLK		(1 << 22)
 
 /* gamt regs */
 #define GEN8_L3_LRA_1_GPGPU _MMIO(0x4dd4)

drivers/gpu/drm/i915/i915_request.h

@@ -28,13 +28,16 @@
 #include <linux/dma-fence.h>
 
 #include "i915_gem.h"
+#include "i915_scheduler.h"
 #include "i915_sw_fence.h"
+#include "i915_scheduler.h"
 
 #include <uapi/drm/i915_drm.h>
 
 struct drm_file;
 struct drm_i915_gem_object;
 struct i915_request;
+struct i915_timeline;
 
 struct intel_wait {
 	struct rb_node node;
@@ -48,44 +51,6 @@ struct intel_signal_node {
 	struct list_head link;
 };
 
-struct i915_dependency {
-	struct i915_priotree *signaler;
-	struct list_head signal_link;
-	struct list_head wait_link;
-	struct list_head dfs_link;
-	unsigned long flags;
-#define I915_DEPENDENCY_ALLOC BIT(0)
-};
-
-/*
- * "People assume that time is a strict progression of cause to effect, but
- * actually, from a nonlinear, non-subjective viewpoint, it's more like a big
- * ball of wibbly-wobbly, timey-wimey ... stuff." -The Doctor, 2015
- *
- * Requests exist in a complex web of interdependencies. Each request
- * has to wait for some other request to complete before it is ready to be run
- * (e.g. we have to wait until the pixels have been rendering into a texture
- * before we can copy from it). We track the readiness of a request in terms
- * of fences, but we also need to keep the dependency tree for the lifetime
- * of the request (beyond the life of an individual fence). We use the tree
- * at various points to reorder the requests whilst keeping the requests
- * in order with respect to their various dependencies.
- */
-struct i915_priotree {
-	struct list_head signalers_list; /* those before us, we depend upon */
-	struct list_head waiters_list; /* those after us, they depend upon us */
-	struct list_head link;
-	int priority;
-};
-
-enum {
-	I915_PRIORITY_MIN = I915_CONTEXT_MIN_USER_PRIORITY - 1,
-	I915_PRIORITY_NORMAL = I915_CONTEXT_DEFAULT_PRIORITY,
-	I915_PRIORITY_MAX = I915_CONTEXT_MAX_USER_PRIORITY + 1,
-
-	I915_PRIORITY_INVALID = INT_MIN
-};
-
 struct i915_capture_list {
 	struct i915_capture_list *next;
 	struct i915_vma *vma;
@@ -131,7 +96,7 @@ struct i915_request {
 	struct i915_gem_context *ctx;
 	struct intel_engine_cs *engine;
 	struct intel_ring *ring;
-	struct intel_timeline *timeline;
+	struct i915_timeline *timeline;
 	struct intel_signal_node signaling;
 
 	/*
@@ -154,7 +119,7 @@ struct i915_request {
 	 * to retirement), i.e. bidirectional dependency information for the
 	 * request not tied to individual fences.
 	 */
-	struct i915_priotree priotree;
+	struct i915_sched_node sched;
 	struct i915_dependency dep;
 
 	/**
@@ -343,10 +308,10 @@ static inline bool i915_request_started(const struct i915_request *rq)
 				 seqno - 1);
 }
 
-static inline bool i915_priotree_signaled(const struct i915_priotree *pt)
+static inline bool i915_sched_node_signaled(const struct i915_sched_node *node)
 {
 	const struct i915_request *rq =
-		container_of(pt, const struct i915_request, priotree);
+		container_of(node, const struct i915_request, sched);
 
 	return i915_request_completed(rq);
 }

drivers/gpu/drm/i915/i915_scheduler.h

+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2018 Intel Corporation
+ */
+
+#ifndef _I915_SCHEDULER_H_
+#define _I915_SCHEDULER_H_
+
+#include <linux/bitops.h>
+
+#include <uapi/drm/i915_drm.h>
+
+enum {
+	I915_PRIORITY_MIN = I915_CONTEXT_MIN_USER_PRIORITY - 1,
+	I915_PRIORITY_NORMAL = I915_CONTEXT_DEFAULT_PRIORITY,
+	I915_PRIORITY_MAX = I915_CONTEXT_MAX_USER_PRIORITY + 1,
+
+	I915_PRIORITY_INVALID = INT_MIN
+};
+
+struct i915_sched_attr {
+	/**
+	 * @priority: execution and service priority
+	 *
+	 * All clients are equal, but some are more equal than others!
+	 *
+	 * Requests from a context with a greater (more positive) value of
+	 * @priority will be executed before those with a lower @priority
+	 * value, forming a simple QoS.
+	 *
+	 * The &drm_i915_private.kernel_context is assigned the lowest priority.
+	 */
+	int priority;
+};
+
+/*
+ * "People assume that time is a strict progression of cause to effect, but
+ * actually, from a nonlinear, non-subjective viewpoint, it's more like a big
+ * ball of wibbly-wobbly, timey-wimey ... stuff." -The Doctor, 2015
+ *
+ * Requests exist in a complex web of interdependencies. Each request
+ * has to wait for some other request to complete before it is ready to be run
+ * (e.g. we have to wait until the pixels have been rendering into a texture
+ * before we can copy from it). We track the readiness of a request in terms
+ * of fences, but we also need to keep the dependency tree for the lifetime
+ * of the request (beyond the life of an individual fence). We use the tree
+ * at various points to reorder the requests whilst keeping the requests
+ * in order with respect to their various dependencies.
+ *
+ * There is no active component to the "scheduler". As we know the dependency
+ * DAG of each request, we are able to insert it into a sorted queue when it
+ * is ready, and are able to reorder its portion of the graph to accommodate
+ * dynamic priority changes.
+ */
+struct i915_sched_node {
+	struct list_head signalers_list; /* those before us, we depend upon */
+	struct list_head waiters_list; /* those after us, they depend upon us */
+	struct list_head link;
+	struct i915_sched_attr attr;
+};
+
+struct i915_dependency {
+	struct i915_sched_node *signaler;
+	struct list_head signal_link;
+	struct list_head wait_link;
+	struct list_head dfs_link;
+	unsigned long flags;
+#define I915_DEPENDENCY_ALLOC BIT(0)
+};
+
+#endif /* _I915_SCHEDULER_H_ */

drivers/gpu/drm/i915/i915_timeline.c

+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2016-2018 Intel Corporation
+ */
+
+#include "i915_drv.h"
+
+#include "i915_timeline.h"
+#include "i915_syncmap.h"
+
+void i915_timeline_init(struct drm_i915_private *i915,
+			struct i915_timeline *timeline,
+			const char *name)
+{
+	lockdep_assert_held(&i915->drm.struct_mutex);
+
+	/*
+	 * Ideally we want a set of engines on a single leaf as we expect
+	 * to mostly be tracking synchronisation between engines. It is not
+	 * a huge issue if this is not the case, but we may want to mitigate
+	 * any page crossing penalties if they become an issue.
+	 */
+	BUILD_BUG_ON(KSYNCMAP < I915_NUM_ENGINES);
+
+	timeline->name = name;
+
+	list_add(&timeline->link, &i915->gt.timelines);
+
+	/* Called during early_init before we know how many engines there are */
+
+	timeline->fence_context = dma_fence_context_alloc(1);
+
+	spin_lock_init(&timeline->lock);
+
+	init_request_active(&timeline->last_request, NULL);
+	INIT_LIST_HEAD(&timeline->requests);
+
+	i915_syncmap_init(&timeline->sync);
+}
+
+/**
+ * i915_timelines_park - called when the driver idles
+ * @i915: the drm_i915_private device
+ *
+ * When the driver is completely idle, we know that all of our sync points
+ * have been signaled and our tracking is then entirely redundant. Any request
+ * to wait upon an older sync point will be completed instantly as we know
+ * the fence is signaled and therefore we will not even look them up in the
+ * sync point map.
+ */
+void i915_timelines_park(struct drm_i915_private *i915)
+{
+	struct i915_timeline *timeline;
+
+	lockdep_assert_held(&i915->drm.struct_mutex);
+
+	list_for_each_entry(timeline, &i915->gt.timelines, link) {
+		/*
+		 * All known fences are completed so we can scrap
+		 * the current sync point tracking and start afresh,
+		 * any attempt to wait upon a previous sync point
+		 * will be skipped as the fence was signaled.
+		 */
+		i915_syncmap_free(&timeline->sync);
+	}
+}
+
+void i915_timeline_fini(struct i915_timeline *timeline)
+{
+	GEM_BUG_ON(!list_empty(&timeline->requests));
+
+	i915_syncmap_free(&timeline->sync);
+
+	list_del(&timeline->link);
+}
+
+struct i915_timeline *
+i915_timeline_create(struct drm_i915_private *i915, const char *name)
+{
+	struct i915_timeline *timeline;
+
+	timeline = kzalloc(sizeof(*timeline), GFP_KERNEL);
+	if (!timeline)
+		return ERR_PTR(-ENOMEM);
+
+	i915_timeline_init(i915, timeline, name);
+	kref_init(&timeline->kref);
+
+	return timeline;
+}
+
+void __i915_timeline_free(struct kref *kref)
+{
+	struct i915_timeline *timeline =
+		container_of(kref, typeof(*timeline), kref);
+
+	i915_timeline_fini(timeline);
+	kfree(timeline);
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/mock_timeline.c"
+#include "selftests/i915_timeline.c"
+#endif

drivers/gpu/drm/i915/i915_gem_timeline.h → drivers/gpu/drm/i915/i915_timeline.h

@@ -22,27 +22,20 @@
  *
  */
 
-#ifndef I915_GEM_TIMELINE_H
-#define I915_GEM_TIMELINE_H
+#ifndef I915_TIMELINE_H
+#define I915_TIMELINE_H
 
 #include <linux/list.h>
+#include <linux/kref.h>
 
 #include "i915_request.h"
 #include "i915_syncmap.h"
 #include "i915_utils.h"
 
-struct i915_gem_timeline;
-
-struct intel_timeline {
+struct i915_timeline {
 	u64 fence_context;
 	u32 seqno;
 
-	/**
-	 * Count of outstanding requests, from the time they are constructed
-	 * to the moment they are retired. Loosely coupled to hardware.
-	 */
-	u32 inflight_seqnos;
-
 	spinlock_t lock;
 
 	/**
@@ -77,47 +70,57 @@ struct intel_timeline {
 	 */
 	u32 global_sync[I915_NUM_ENGINES];
 
-	struct i915_gem_timeline *common;
-};
-
-struct i915_gem_timeline {
 	struct list_head link;
-
-	struct drm_i915_private *i915;
 	const char *name;
 
-	struct intel_timeline engine[I915_NUM_ENGINES];
+	struct kref kref;
 };
 
-int i915_gem_timeline_init(struct drm_i915_private *i915,
-			   struct i915_gem_timeline *tl,
-			   const char *name);
-int i915_gem_timeline_init__global(struct drm_i915_private *i915);
-void i915_gem_timelines_park(struct drm_i915_private *i915);
-void i915_gem_timeline_fini(struct i915_gem_timeline *tl);
+void i915_timeline_init(struct drm_i915_private *i915,
+			struct i915_timeline *tl,
+			const char *name);
+void i915_timeline_fini(struct i915_timeline *tl);
+
+struct i915_timeline *
+i915_timeline_create(struct drm_i915_private *i915, const char *name);
 
-static inline int __intel_timeline_sync_set(struct intel_timeline *tl,
-					    u64 context, u32 seqno)
+static inline struct i915_timeline *
+i915_timeline_get(struct i915_timeline *timeline)
+{
+	kref_get(&timeline->kref);
+	return timeline;
+}
+
+void __i915_timeline_free(struct kref *kref);
+static inline void i915_timeline_put(struct i915_timeline *timeline)
+{
+	kref_put(&timeline->kref, __i915_timeline_free);
+}
+
+static inline int __i915_timeline_sync_set(struct i915_timeline *tl,
+					   u64 context, u32 seqno)
 {
 	return i915_syncmap_set(&tl->sync, context, seqno);
 }
 
-static inline int intel_timeline_sync_set(struct intel_timeline *tl,
-					  const struct dma_fence *fence)
+static inline int i915_timeline_sync_set(struct i915_timeline *tl,
+					 const struct dma_fence *fence)
 {
-	return __intel_timeline_sync_set(tl, fence->context, fence->seqno);
+	return __i915_timeline_sync_set(tl, fence->context, fence->seqno);
 }
 
-static inline bool __intel_timeline_sync_is_later(struct intel_timeline *tl,
-						  u64 context, u32 seqno)
+static inline bool __i915_timeline_sync_is_later(struct i915_timeline *tl,
+						 u64 context, u32 seqno)
 {
 	return i915_syncmap_is_later(&tl->sync, context, seqno);
 }
 
-static inline bool intel_timeline_sync_is_later(struct intel_timeline *tl,
-						const struct dma_fence *fence)
+static inline bool i915_timeline_sync_is_later(struct i915_timeline *tl,
+					       const struct dma_fence *fence)
 {
-	return __intel_timeline_sync_is_later(tl, fence->context, fence->seqno);
+	return __i915_timeline_sync_is_later(tl, fence->context, fence->seqno);
 }
 
+void i915_timelines_park(struct drm_i915_private *i915);
+
 #endif

drivers/gpu/drm/i915/i915_trace.h

@@ -679,45 +679,68 @@ DEFINE_EVENT(i915_request, i915_request_execute,
 	     TP_ARGS(rq)
 );
 
-DECLARE_EVENT_CLASS(i915_request_hw,
-		    TP_PROTO(struct i915_request *rq, unsigned int port),
-		    TP_ARGS(rq, port),
-
-		    TP_STRUCT__entry(
-				     __field(u32, dev)
-				     __field(u32, hw_id)
-				     __field(u32, ring)
-				     __field(u32, ctx)
-				     __field(u32, seqno)
-				     __field(u32, global_seqno)
-				     __field(u32, port)
-				    ),
-
-		    TP_fast_assign(
-				   __entry->dev = rq->i915->drm.primary->index;
-				   __entry->hw_id = rq->ctx->hw_id;
-				   __entry->ring = rq->engine->id;
-				   __entry->ctx = rq->fence.context;
-				   __entry->seqno = rq->fence.seqno;
-				   __entry->global_seqno = rq->global_seqno;
-				   __entry->port = port;
-				  ),
-
-		    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, global=%u, port=%u",
-			      __entry->dev, __entry->hw_id, __entry->ring,
-			      __entry->ctx, __entry->seqno,
-			      __entry->global_seqno, __entry->port)
-);
+TRACE_EVENT(i915_request_in,
+	    TP_PROTO(struct i915_request *rq, unsigned int port),
+	    TP_ARGS(rq, port),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     __field(u32, hw_id)
+			     __field(u32, ring)
+			     __field(u32, ctx)
+			     __field(u32, seqno)
+			     __field(u32, global_seqno)
+			     __field(u32, port)
+			     __field(u32, prio)
+			    ),
+
+	    TP_fast_assign(
+			   __entry->dev = rq->i915->drm.primary->index;
+			   __entry->hw_id = rq->ctx->hw_id;
+			   __entry->ring = rq->engine->id;
+			   __entry->ctx = rq->fence.context;
+			   __entry->seqno = rq->fence.seqno;
+			   __entry->global_seqno = rq->global_seqno;
+			   __entry->prio = rq->sched.attr.priority;
+			   __entry->port = port;
+			   ),
 
-DEFINE_EVENT(i915_request_hw, i915_request_in,
-	     TP_PROTO(struct i915_request *rq, unsigned int port),
-	     TP_ARGS(rq, port)
+	    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, prio=%u, global=%u, port=%u",
+		      __entry->dev, __entry->hw_id, __entry->ring, __entry->ctx,
+		      __entry->seqno, __entry->prio, __entry->global_seqno,
+		      __entry->port)
 );
 
-DEFINE_EVENT(i915_request, i915_request_out,
-	     TP_PROTO(struct i915_request *rq),
-	     TP_ARGS(rq)
+TRACE_EVENT(i915_request_out,
+	    TP_PROTO(struct i915_request *rq),
+	    TP_ARGS(rq),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     __field(u32, hw_id)
+			     __field(u32, ring)
+			     __field(u32, ctx)
+			     __field(u32, seqno)
+			     __field(u32, global_seqno)
+			     __field(u32, completed)
+			    ),
+
+	    TP_fast_assign(
+			   __entry->dev = rq->i915->drm.primary->index;
+			   __entry->hw_id = rq->ctx->hw_id;
+			   __entry->ring = rq->engine->id;
+			   __entry->ctx = rq->fence.context;
+			   __entry->seqno = rq->fence.seqno;
+			   __entry->global_seqno = rq->global_seqno;
+			   __entry->completed = i915_request_completed(rq);
+			   ),
+
+		    TP_printk("dev=%u, hw_id=%u, ring=%u, ctx=%u, seqno=%u, global=%u, completed?=%u",
+			      __entry->dev, __entry->hw_id, __entry->ring,
+			      __entry->ctx, __entry->seqno,
+			      __entry->global_seqno, __entry->completed)
 );
+
 #else
 #if !defined(TRACE_HEADER_MULTI_READ)
 static inline void
@@ -811,42 +834,6 @@ DEFINE_EVENT(i915_request, i915_request_wait_end,
 	    TP_ARGS(rq)
 );
 
-TRACE_EVENT(i915_flip_request,
-	    TP_PROTO(int plane, struct drm_i915_gem_object *obj),
-
-	    TP_ARGS(plane, obj),
-
-	    TP_STRUCT__entry(
-		    __field(int, plane)
-		    __field(struct drm_i915_gem_object *, obj)
-		    ),
-
-	    TP_fast_assign(
-		    __entry->plane = plane;
-		    __entry->obj = obj;
-		    ),
-
-	    TP_printk("plane=%d, obj=%p", __entry->plane, __entry->obj)
-);
-
-TRACE_EVENT(i915_flip_complete,
-	    TP_PROTO(int plane, struct drm_i915_gem_object *obj),
-
-	    TP_ARGS(plane, obj),
-
-	    TP_STRUCT__entry(
-		    __field(int, plane)
-		    __field(struct drm_i915_gem_object *, obj)
-		    ),
-
-	    TP_fast_assign(
-		    __entry->plane = plane;
-		    __entry->obj = obj;
-		    ),
-
-	    TP_printk("plane=%d, obj=%p", __entry->plane, __entry->obj)
-);
-
 TRACE_EVENT_CONDITION(i915_reg_rw,
 	TP_PROTO(bool write, i915_reg_t reg, u64 val, int len, bool trace),
 

drivers/gpu/drm/i915/i915_utils.h

@@ -120,6 +120,12 @@ static inline u64 ptr_to_u64(const void *ptr)
 
 #include <linux/list.h>
 
+static inline int list_is_first(const struct list_head *list,
+				const struct list_head *head)
+{
+	return head->next == list;
+}
+
 static inline void __list_del_many(struct list_head *head,
 				   struct list_head *first)
 {

drivers/gpu/drm/i915/i915_vma.c

@@ -46,8 +46,6 @@ i915_vma_retire(struct i915_gem_active *active, struct i915_request *rq)
 
 	GEM_BUG_ON(!drm_mm_node_allocated(&vma->node));
 	list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
-	if (unlikely(i915_vma_is_closed(vma) && !i915_vma_is_pinned(vma)))
-		WARN_ON(i915_vma_unbind(vma));
 
 	GEM_BUG_ON(!i915_gem_object_is_active(obj));
 	if (--obj->active_count)
@@ -232,7 +230,6 @@ i915_vma_instance(struct drm_i915_gem_object *obj,
 	if (!vma)
 		vma = vma_create(obj, vm, view);
 
-	GEM_BUG_ON(!IS_ERR(vma) && i915_vma_is_closed(vma));
 	GEM_BUG_ON(!IS_ERR(vma) && i915_vma_compare(vma, vm, view));
 	GEM_BUG_ON(!IS_ERR(vma) && vma_lookup(obj, vm, view) != vma);
 	return vma;
@@ -684,13 +681,43 @@ int __i915_vma_do_pin(struct i915_vma *vma,
 	return ret;
 }
 
-static void i915_vma_destroy(struct i915_vma *vma)
+void i915_vma_close(struct i915_vma *vma)
+{
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+
+	GEM_BUG_ON(i915_vma_is_closed(vma));
+	vma->flags |= I915_VMA_CLOSED;
+
+	/*
+	 * We defer actually closing, unbinding and destroying the VMA until
+	 * the next idle point, or if the object is freed in the meantime. By
+	 * postponing the unbind, we allow for it to be resurrected by the
+	 * client, avoiding the work required to rebind the VMA. This is
+	 * advantageous for DRI, where the client/server pass objects
+	 * between themselves, temporarily opening a local VMA to the
+	 * object, and then closing it again. The same object is then reused
+	 * on the next frame (or two, depending on the depth of the swap queue)
+	 * causing us to rebind the VMA once more. This ends up being a lot
+	 * of wasted work for the steady state.
+	 */
+	list_add_tail(&vma->closed_link, &vma->vm->i915->gt.closed_vma);
+}
+
+void i915_vma_reopen(struct i915_vma *vma)
+{
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
+
+	if (vma->flags & I915_VMA_CLOSED) {
+		vma->flags &= ~I915_VMA_CLOSED;
+		list_del(&vma->closed_link);
+	}
+}
+
+static void __i915_vma_destroy(struct i915_vma *vma)
 {
 	int i;
 
 	GEM_BUG_ON(vma->node.allocated);
-	GEM_BUG_ON(i915_vma_is_active(vma));
-	GEM_BUG_ON(!i915_vma_is_closed(vma));
 	GEM_BUG_ON(vma->fence);
 
 	for (i = 0; i < ARRAY_SIZE(vma->last_read); i++)
@@ -699,6 +726,7 @@ static void i915_vma_destroy(struct i915_vma *vma)
 
 	list_del(&vma->obj_link);
 	list_del(&vma->vm_link);
+	rb_erase(&vma->obj_node, &vma->obj->vma_tree);
 
 	if (!i915_vma_is_ggtt(vma))
 		i915_ppgtt_put(i915_vm_to_ppgtt(vma->vm));
@@ -706,15 +734,30 @@ static void i915_vma_destroy(struct i915_vma *vma)
 	kmem_cache_free(to_i915(vma->obj->base.dev)->vmas, vma);
 }
 
-void i915_vma_close(struct i915_vma *vma)
+void i915_vma_destroy(struct i915_vma *vma)
 {
-	GEM_BUG_ON(i915_vma_is_closed(vma));
-	vma->flags |= I915_VMA_CLOSED;
+	lockdep_assert_held(&vma->vm->i915->drm.struct_mutex);
 
-	rb_erase(&vma->obj_node, &vma->obj->vma_tree);
+	GEM_BUG_ON(i915_vma_is_active(vma));
+	GEM_BUG_ON(i915_vma_is_pinned(vma));
+
+	if (i915_vma_is_closed(vma))
+		list_del(&vma->closed_link);
+
+	WARN_ON(i915_vma_unbind(vma));
+	__i915_vma_destroy(vma);
+}
+
+void i915_vma_parked(struct drm_i915_private *i915)
+{
+	struct i915_vma *vma, *next;
 
-	if (!i915_vma_is_active(vma) && !i915_vma_is_pinned(vma))
-		WARN_ON(i915_vma_unbind(vma));
+	list_for_each_entry_safe(vma, next, &i915->gt.closed_vma, closed_link) {
+		GEM_BUG_ON(!i915_vma_is_closed(vma));
+		i915_vma_destroy(vma);
+	}
+
+	GEM_BUG_ON(!list_empty(&i915->gt.closed_vma));
 }
 
 static void __i915_vma_iounmap(struct i915_vma *vma)
@@ -804,7 +847,7 @@ int i915_vma_unbind(struct i915_vma *vma)
 		return -EBUSY;
 
 	if (!drm_mm_node_allocated(&vma->node))
-		goto destroy;
+		return 0;
 
 	GEM_BUG_ON(obj->bind_count == 0);
 	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
@@ -841,10 +884,6 @@ int i915_vma_unbind(struct i915_vma *vma)
 
 	i915_vma_remove(vma);
 
-destroy:
-	if (unlikely(i915_vma_is_closed(vma)))
-		i915_vma_destroy(vma);
-
 	return 0;
 }
 

drivers/gpu/drm/i915/i915_vma.h

@@ -119,6 +119,8 @@ struct i915_vma {
 	/** This vma's place in the eviction list */
 	struct list_head evict_link;
 
+	struct list_head closed_link;
+
 	/**
 	 * Used for performing relocations during execbuffer insertion.
 	 */
@@ -285,6 +287,8 @@ void i915_vma_revoke_mmap(struct i915_vma *vma);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
 void i915_vma_unlink_ctx(struct i915_vma *vma);
 void i915_vma_close(struct i915_vma *vma);
+void i915_vma_reopen(struct i915_vma *vma);
+void i915_vma_destroy(struct i915_vma *vma);
 
 int __i915_vma_do_pin(struct i915_vma *vma,
 		      u64 size, u64 alignment, u64 flags);
@@ -408,6 +412,8 @@ i915_vma_unpin_fence(struct i915_vma *vma)
 		__i915_vma_unpin_fence(vma);
 }
 
+void i915_vma_parked(struct drm_i915_private *i915);
+
 #define for_each_until(cond) if (cond) break; else
 
 /**

drivers/gpu/drm/i915/intel_atomic_plane.c

@@ -183,11 +183,16 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 	}
 
 	/* FIXME pre-g4x don't work like this */
-	if (intel_state->base.visible)
+	if (state->visible)
 		crtc_state->active_planes |= BIT(intel_plane->id);
 	else
 		crtc_state->active_planes &= ~BIT(intel_plane->id);
 
+	if (state->visible && state->fb->format->format == DRM_FORMAT_NV12)
+		crtc_state->nv12_planes |= BIT(intel_plane->id);
+	else
+		crtc_state->nv12_planes &= ~BIT(intel_plane->id);
+
 	return intel_plane_atomic_calc_changes(old_crtc_state,
 					       &crtc_state->base,
 					       old_plane_state,

drivers/gpu/drm/i915/intel_bios.c

@@ -530,6 +530,7 @@ parse_driver_features(struct drm_i915_private *dev_priv,
 	 */
 	if (!driver->drrs_enabled)
 		dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
+	dev_priv->vbt.psr.enable = driver->psr_enabled;
 }
 
 static void

drivers/gpu/drm/i915/intel_breadcrumbs.c

@@ -82,7 +82,7 @@ static unsigned long wait_timeout(void)
 
 static noinline void missed_breadcrumb(struct intel_engine_cs *engine)
 {
-	if (drm_debug & DRM_UT_DRIVER) {
+	if (GEM_SHOW_DEBUG()) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
 		intel_engine_dump(engine, &p,
@@ -130,11 +130,12 @@ static void intel_breadcrumbs_hangcheck(struct timer_list *t)
 
 static void intel_breadcrumbs_fake_irq(struct timer_list *t)
 {
-	struct intel_engine_cs *engine = from_timer(engine, t,
-						    breadcrumbs.fake_irq);
+	struct intel_engine_cs *engine =
+		from_timer(engine, t, breadcrumbs.fake_irq);
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
-	/* The timer persists in case we cannot enable interrupts,
+	/*
+	 * The timer persists in case we cannot enable interrupts,
 	 * or if we have previously seen seqno/interrupt incoherency
 	 * ("missed interrupt" syndrome, better known as a "missed breadcrumb").
 	 * Here the worker will wake up every jiffie in order to kick the
@@ -148,6 +149,12 @@ static void intel_breadcrumbs_fake_irq(struct timer_list *t)
 	if (!b->irq_armed)
 		return;
 
+	/* If the user has disabled the fake-irq, restore the hangchecking */
+	if (!test_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings)) {
+		mod_timer(&b->hangcheck, wait_timeout());
+		return;
+	}
+
 	mod_timer(&b->fake_irq, jiffies + 1);
 }
 
@@ -831,8 +838,8 @@ static void cancel_fake_irq(struct intel_engine_cs *engine)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
+	del_timer_sync(&b->fake_irq); /* may queue b->hangcheck */
 	del_timer_sync(&b->hangcheck);
-	del_timer_sync(&b->fake_irq);
 	clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
 }
 
@@ -840,15 +847,22 @@ void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 
-	cancel_fake_irq(engine);
 	spin_lock_irq(&b->irq_lock);
 
+	/*
+	 * Leave the fake_irq timer enabled (if it is running), but clear the
+	 * bit so that it turns itself off on its next wake up and goes back
+	 * to the long hangcheck interval if still required.
+	 */
+	clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
+
 	if (b->irq_enabled)
 		irq_enable(engine);
 	else
 		irq_disable(engine);
 
-	/* We set the IRQ_BREADCRUMB bit when we enable the irq presuming the
+	/*
+	 * We set the IRQ_BREADCRUMB bit when we enable the irq presuming the
 	 * GPU is active and may have already executed the MI_USER_INTERRUPT
 	 * before the CPU is ready to receive. However, the engine is currently
 	 * idle (we haven't started it yet), there is no possibility for a
@@ -857,9 +871,6 @@ void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine)
 	 */
 	clear_bit(ENGINE_IRQ_BREADCRUMB, &engine->irq_posted);
 
-	if (b->irq_armed)
-		enable_fake_irq(b);
-
 	spin_unlock_irq(&b->irq_lock);
 }
 

drivers/gpu/drm/i915/intel_cdclk.c

@@ -2302,9 +2302,44 @@ static int bdw_modeset_calc_cdclk(struct drm_atomic_state *state)
 	return 0;
 }
 
+static int skl_dpll0_vco(struct intel_atomic_state *intel_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(intel_state->base.dev);
+	struct intel_crtc *crtc;
+	struct intel_crtc_state *crtc_state;
+	int vco, i;
+
+	vco = intel_state->cdclk.logical.vco;
+	if (!vco)
+		vco = dev_priv->skl_preferred_vco_freq;
+
+	for_each_new_intel_crtc_in_state(intel_state, crtc, crtc_state, i) {
+		if (!crtc_state->base.enable)
+			continue;
+
+		if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_EDP))
+			continue;
+
+		/*
+		 * DPLL0 VCO may need to be adjusted to get the correct
+		 * clock for eDP. This will affect cdclk as well.
+		 */
+		switch (crtc_state->port_clock / 2) {
+		case 108000:
+		case 216000:
+			vco = 8640000;
+			break;
+		default:
+			vco = 8100000;
+			break;
+		}
+	}
+
+	return vco;
+}
+
 static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(state->dev);
 	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
 	int min_cdclk, cdclk, vco;
 
@@ -2312,9 +2347,7 @@ static int skl_modeset_calc_cdclk(struct drm_atomic_state *state)
 	if (min_cdclk < 0)
 		return min_cdclk;
 
-	vco = intel_state->cdclk.logical.vco;
-	if (!vco)
-		vco = dev_priv->skl_preferred_vco_freq;
+	vco = skl_dpll0_vco(intel_state);
 
 	/*
 	 * FIXME should also account for plane ratio

drivers/gpu/drm/i915/intel_csr.c

@@ -298,7 +298,10 @@ static uint32_t *parse_csr_fw(struct drm_i915_private *dev_priv,
 
 	csr->version = css_header->version;
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (csr->fw_path == i915_modparams.dmc_firmware_path) {
+		/* Bypass version check for firmware override. */
+		required_version = csr->version;
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		required_version = CNL_CSR_VERSION_REQUIRED;
 	} else if (IS_GEMINILAKE(dev_priv)) {
 		required_version = GLK_CSR_VERSION_REQUIRED;
@@ -453,7 +456,9 @@ void intel_csr_ucode_init(struct drm_i915_private *dev_priv)
 	if (!HAS_CSR(dev_priv))
 		return;
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (i915_modparams.dmc_firmware_path)
+		csr->fw_path = i915_modparams.dmc_firmware_path;
+	else if (IS_CANNONLAKE(dev_priv))
 		csr->fw_path = I915_CSR_CNL;
 	else if (IS_GEMINILAKE(dev_priv))
 		csr->fw_path = I915_CSR_GLK;

drivers/gpu/drm/i915/intel_ddi.c

@@ -870,6 +870,45 @@ cnl_get_buf_trans_edp(struct drm_i915_private *dev_priv, int *n_entries)
 	}
 }
 
+static const struct icl_combo_phy_ddi_buf_trans *
+icl_get_combo_buf_trans(struct drm_i915_private *dev_priv, enum port port,
+			int type, int *n_entries)
+{
+	u32 voltage = I915_READ(ICL_PORT_COMP_DW3(port)) & VOLTAGE_INFO_MASK;
+
+	if (type == INTEL_OUTPUT_EDP && dev_priv->vbt.edp.low_vswing) {
+		switch (voltage) {
+		case VOLTAGE_INFO_0_85V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_0_85V);
+			return icl_combo_phy_ddi_translations_edp_0_85V;
+		case VOLTAGE_INFO_0_95V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_0_95V);
+			return icl_combo_phy_ddi_translations_edp_0_95V;
+		case VOLTAGE_INFO_1_05V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_edp_1_05V);
+			return icl_combo_phy_ddi_translations_edp_1_05V;
+		default:
+			MISSING_CASE(voltage);
+			return NULL;
+		}
+	} else {
+		switch (voltage) {
+		case VOLTAGE_INFO_0_85V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_0_85V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_0_85V;
+		case VOLTAGE_INFO_0_95V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_0_95V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_0_95V;
+		case VOLTAGE_INFO_1_05V:
+			*n_entries = ARRAY_SIZE(icl_combo_phy_ddi_translations_dp_hdmi_1_05V);
+			return icl_combo_phy_ddi_translations_dp_hdmi_1_05V;
+		default:
+			MISSING_CASE(voltage);
+			return NULL;
+		}
+	}
+}
+
 static int intel_ddi_hdmi_level(struct drm_i915_private *dev_priv, enum port port)
 {
 	int n_entries, level, default_entry;
@@ -1013,6 +1052,25 @@ static uint32_t hsw_pll_to_ddi_pll_sel(const struct intel_shared_dpll *pll)
 	}
 }
 
+static uint32_t icl_pll_to_ddi_pll_sel(struct intel_encoder *encoder,
+				       const struct intel_shared_dpll *pll)
+{
+	const enum intel_dpll_id id = pll->info->id;
+
+	switch (id) {
+	default:
+		MISSING_CASE(id);
+	case DPLL_ID_ICL_DPLL0:
+	case DPLL_ID_ICL_DPLL1:
+		return DDI_CLK_SEL_NONE;
+	case DPLL_ID_ICL_MGPLL1:
+	case DPLL_ID_ICL_MGPLL2:
+	case DPLL_ID_ICL_MGPLL3:
+	case DPLL_ID_ICL_MGPLL4:
+		return DDI_CLK_SEL_MG;
+	}
+}
+
 /* Starting with Haswell, different DDI ports can work in FDI mode for
  * connection to the PCH-located connectors. For this, it is necessary to train
  * both the DDI port and PCH receiver for the desired DDI buffer settings.
@@ -2025,7 +2083,13 @@ u8 intel_ddi_dp_voltage_max(struct intel_encoder *encoder)
 	enum port port = encoder->port;
 	int n_entries;
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (IS_ICELAKE(dev_priv)) {
+		if (port == PORT_A || port == PORT_B)
+			icl_get_combo_buf_trans(dev_priv, port, encoder->type,
+						&n_entries);
+		else
+			n_entries = ARRAY_SIZE(icl_mg_phy_ddi_translations);
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		if (encoder->type == INTEL_OUTPUT_EDP)
 			cnl_get_buf_trans_edp(dev_priv, &n_entries);
 		else
@@ -2182,6 +2246,146 @@ static void cnl_ddi_vswing_sequence(struct intel_encoder *encoder,
 	I915_WRITE(CNL_PORT_TX_DW5_GRP(port), val);
 }
 
+static void icl_ddi_combo_vswing_program(struct drm_i915_private *dev_priv,
+					 u32 level, enum port port, int type)
+{
+	const struct icl_combo_phy_ddi_buf_trans *ddi_translations = NULL;
+	u32 n_entries, val;
+	int ln;
+
+	ddi_translations = icl_get_combo_buf_trans(dev_priv, port, type,
+						   &n_entries);
+	if (!ddi_translations)
+		return;
+
+	if (level >= n_entries) {
+		DRM_DEBUG_KMS("DDI translation not found for level %d. Using %d instead.", level, n_entries - 1);
+		level = n_entries - 1;
+	}
+
+	/* Set PORT_TX_DW5 Rterm Sel to 110b. */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val &= ~RTERM_SELECT_MASK;
+	val |= RTERM_SELECT(0x6);
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* Program PORT_TX_DW5 */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	/* Set DisableTap2 and DisableTap3 if MIPI DSI
+	 * Clear DisableTap2 and DisableTap3 for all other Ports
+	 */
+	if (type == INTEL_OUTPUT_DSI) {
+		val |= TAP2_DISABLE;
+		val |= TAP3_DISABLE;
+	} else {
+		val &= ~TAP2_DISABLE;
+		val &= ~TAP3_DISABLE;
+	}
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* Program PORT_TX_DW2 */
+	val = I915_READ(ICL_PORT_TX_DW2_LN0(port));
+	val &= ~(SWING_SEL_LOWER_MASK | SWING_SEL_UPPER_MASK |
+		 RCOMP_SCALAR_MASK);
+	val |= SWING_SEL_UPPER(ddi_translations[level].dw2_swing_select);
+	val |= SWING_SEL_LOWER(ddi_translations[level].dw2_swing_select);
+	/* Program Rcomp scalar for every table entry */
+	val |= RCOMP_SCALAR(ddi_translations[level].dw2_swing_scalar);
+	I915_WRITE(ICL_PORT_TX_DW2_GRP(port), val);
+
+	/* Program PORT_TX_DW4 */
+	/* We cannot write to GRP. It would overwrite individual loadgen. */
+	for (ln = 0; ln <= 3; ln++) {
+		val = I915_READ(ICL_PORT_TX_DW4_LN(port, ln));
+		val &= ~(POST_CURSOR_1_MASK | POST_CURSOR_2_MASK |
+			 CURSOR_COEFF_MASK);
+		val |= ddi_translations[level].dw4_scaling;
+		I915_WRITE(ICL_PORT_TX_DW4_LN(port, ln), val);
+	}
+}
+
+static void icl_combo_phy_ddi_vswing_sequence(struct intel_encoder *encoder,
+					      u32 level,
+					      enum intel_output_type type)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	enum port port = encoder->port;
+	int width = 0;
+	int rate = 0;
+	u32 val;
+	int ln = 0;
+
+	if (type == INTEL_OUTPUT_HDMI) {
+		width = 4;
+		/* Rate is always < than 6GHz for HDMI */
+	} else {
+		struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+
+		width = intel_dp->lane_count;
+		rate = intel_dp->link_rate;
+	}
+
+	/*
+	 * 1. If port type is eDP or DP,
+	 * set PORT_PCS_DW1 cmnkeeper_enable to 1b,
+	 * else clear to 0b.
+	 */
+	val = I915_READ(ICL_PORT_PCS_DW1_LN0(port));
+	if (type == INTEL_OUTPUT_HDMI)
+		val &= ~COMMON_KEEPER_EN;
+	else
+		val |= COMMON_KEEPER_EN;
+	I915_WRITE(ICL_PORT_PCS_DW1_GRP(port), val);
+
+	/* 2. Program loadgen select */
+	/*
+	 * Program PORT_TX_DW4_LN depending on Bit rate and used lanes
+	 * <= 6 GHz and 4 lanes (LN0=0, LN1=1, LN2=1, LN3=1)
+	 * <= 6 GHz and 1,2 lanes (LN0=0, LN1=1, LN2=1, LN3=0)
+	 * > 6 GHz (LN0=0, LN1=0, LN2=0, LN3=0)
+	 */
+	for (ln = 0; ln <= 3; ln++) {
+		val = I915_READ(ICL_PORT_TX_DW4_LN(port, ln));
+		val &= ~LOADGEN_SELECT;
+
+		if ((rate <= 600000 && width == 4 && ln >= 1) ||
+		    (rate <= 600000 && width < 4 && (ln == 1 || ln == 2))) {
+			val |= LOADGEN_SELECT;
+		}
+		I915_WRITE(ICL_PORT_TX_DW4_LN(port, ln), val);
+	}
+
+	/* 3. Set PORT_CL_DW5 SUS Clock Config to 11b */
+	val = I915_READ(ICL_PORT_CL_DW5(port));
+	val |= SUS_CLOCK_CONFIG;
+	I915_WRITE(ICL_PORT_CL_DW5(port), val);
+
+	/* 4. Clear training enable to change swing values */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val &= ~TX_TRAINING_EN;
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+
+	/* 5. Program swing and de-emphasis */
+	icl_ddi_combo_vswing_program(dev_priv, level, port, type);
+
+	/* 6. Set training enable to trigger update */
+	val = I915_READ(ICL_PORT_TX_DW5_LN0(port));
+	val |= TX_TRAINING_EN;
+	I915_WRITE(ICL_PORT_TX_DW5_GRP(port), val);
+}
+
+static void icl_ddi_vswing_sequence(struct intel_encoder *encoder, u32 level,
+				    enum intel_output_type type)
+{
+	enum port port = encoder->port;
+
+	if (port == PORT_A || port == PORT_B)
+		icl_combo_phy_ddi_vswing_sequence(encoder, level, type);
+	else
+		/* Not Implemented Yet */
+		WARN_ON(1);
+}
+
 static uint32_t translate_signal_level(int signal_levels)
 {
 	int i;
@@ -2213,7 +2417,9 @@ u32 bxt_signal_levels(struct intel_dp *intel_dp)
 	struct intel_encoder *encoder = &dport->base;
 	int level = intel_ddi_dp_level(intel_dp);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, encoder->type);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, encoder->type);
 	else
 		bxt_ddi_vswing_sequence(encoder, level, encoder->type);
@@ -2234,6 +2440,69 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp)
 	return DDI_BUF_TRANS_SELECT(level);
 }
 
+void icl_map_plls_to_ports(struct drm_crtc *crtc,
+			   struct intel_crtc_state *crtc_state,
+			   struct drm_atomic_state *old_state)
+{
+	struct intel_shared_dpll *pll = crtc_state->shared_dpll;
+	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
+	struct drm_connector_state *conn_state;
+	struct drm_connector *conn;
+	int i;
+
+	for_each_new_connector_in_state(old_state, conn, conn_state, i) {
+		struct intel_encoder *encoder =
+			to_intel_encoder(conn_state->best_encoder);
+		enum port port = encoder->port;
+		uint32_t val;
+
+		if (conn_state->crtc != crtc)
+			continue;
+
+		mutex_lock(&dev_priv->dpll_lock);
+
+		val = I915_READ(DPCLKA_CFGCR0_ICL);
+		WARN_ON((val & DPCLKA_CFGCR0_DDI_CLK_OFF(port)) == 0);
+
+		if (port == PORT_A || port == PORT_B) {
+			val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port);
+			val |= DPCLKA_CFGCR0_DDI_CLK_SEL(pll->info->id, port);
+			I915_WRITE(DPCLKA_CFGCR0_ICL, val);
+			POSTING_READ(DPCLKA_CFGCR0_ICL);
+		}
+
+		val &= ~DPCLKA_CFGCR0_DDI_CLK_OFF(port);
+		I915_WRITE(DPCLKA_CFGCR0_ICL, val);
+
+		mutex_unlock(&dev_priv->dpll_lock);
+	}
+}
+
+void icl_unmap_plls_to_ports(struct drm_crtc *crtc,
+			     struct intel_crtc_state *crtc_state,
+			     struct drm_atomic_state *old_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
+	struct drm_connector_state *old_conn_state;
+	struct drm_connector *conn;
+	int i;
+
+	for_each_old_connector_in_state(old_state, conn, old_conn_state, i) {
+		struct intel_encoder *encoder =
+			to_intel_encoder(old_conn_state->best_encoder);
+		enum port port = encoder->port;
+
+		if (old_conn_state->crtc != crtc)
+			continue;
+
+		mutex_lock(&dev_priv->dpll_lock);
+		I915_WRITE(DPCLKA_CFGCR0_ICL,
+			   I915_READ(DPCLKA_CFGCR0_ICL) |
+			   DPCLKA_CFGCR0_DDI_CLK_OFF(port));
+		mutex_unlock(&dev_priv->dpll_lock);
+	}
+}
+
 static void intel_ddi_clk_select(struct intel_encoder *encoder,
 				 const struct intel_shared_dpll *pll)
 {
@@ -2246,7 +2515,11 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 
 	mutex_lock(&dev_priv->dpll_lock);
 
-	if (IS_CANNONLAKE(dev_priv)) {
+	if (IS_ICELAKE(dev_priv)) {
+		if (port >= PORT_C)
+			I915_WRITE(DDI_CLK_SEL(port),
+				   icl_pll_to_ddi_pll_sel(encoder, pll));
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		/* Configure DPCLKA_CFGCR0 to map the DPLL to the DDI. */
 		val = I915_READ(DPCLKA_CFGCR0);
 		val &= ~DPCLKA_CFGCR0_DDI_CLK_SEL_MASK(port);
@@ -2284,14 +2557,18 @@ static void intel_ddi_clk_disable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	enum port port = encoder->port;
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv)) {
+		if (port >= PORT_C)
+			I915_WRITE(DDI_CLK_SEL(port), DDI_CLK_SEL_NONE);
+	} else if (IS_CANNONLAKE(dev_priv)) {
 		I915_WRITE(DPCLKA_CFGCR0, I915_READ(DPCLKA_CFGCR0) |
 			   DPCLKA_CFGCR0_DDI_CLK_OFF(port));
-	else if (IS_GEN9_BC(dev_priv))
+	} else if (IS_GEN9_BC(dev_priv)) {
 		I915_WRITE(DPLL_CTRL2, I915_READ(DPLL_CTRL2) |
 			   DPLL_CTRL2_DDI_CLK_OFF(port));
-	else if (INTEL_GEN(dev_priv) < 9)
+	} else if (INTEL_GEN(dev_priv) < 9) {
 		I915_WRITE(PORT_CLK_SEL(port), PORT_CLK_SEL_NONE);
+	}
 }
 
 static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
@@ -2316,7 +2593,9 @@ static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
 
 	intel_display_power_get(dev_priv, dig_port->ddi_io_power_domain);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, encoder->type);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, encoder->type);
 	else if (IS_GEN9_LP(dev_priv))
 		bxt_ddi_vswing_sequence(encoder, level, encoder->type);
@@ -2347,7 +2626,9 @@ static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
 
 	intel_display_power_get(dev_priv, dig_port->ddi_io_power_domain);
 
-	if (IS_CANNONLAKE(dev_priv))
+	if (IS_ICELAKE(dev_priv))
+		icl_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);
+	else if (IS_CANNONLAKE(dev_priv))
 		cnl_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);
 	else if (IS_GEN9_LP(dev_priv))
 		bxt_ddi_vswing_sequence(encoder, level, INTEL_OUTPUT_HDMI);

drivers/gpu/drm/i915/intel_device_info.c

@@ -848,7 +848,7 @@ void intel_device_info_runtime_init(struct intel_device_info *info)
 		gen9_sseu_info_init(dev_priv);
 	else if (INTEL_GEN(dev_priv) == 10)
 		gen10_sseu_info_init(dev_priv);
-	else if (INTEL_INFO(dev_priv)->gen >= 11)
+	else if (INTEL_GEN(dev_priv) >= 11)
 		gen11_sseu_info_init(dev_priv);
 
 	/* Initialize command stream timestamp frequency */

drivers/gpu/drm/i915/intel_display.h

@@ -218,6 +218,10 @@ struct intel_link_m_n {
 	for ((__p) = 0; (__p) < INTEL_INFO(__dev_priv)->num_pipes; (__p)++) \
 		for_each_if((__mask) & BIT(__p))
 
+#define for_each_cpu_transcoder_masked(__dev_priv, __t, __mask) \
+	for ((__t) = 0; (__t) < I915_MAX_TRANSCODERS; (__t)++)	\
+		for_each_if ((__mask) & (1 << (__t)))
+
 #define for_each_universal_plane(__dev_priv, __pipe, __p)		\
 	for ((__p) = 0;							\
 	     (__p) < INTEL_INFO(__dev_priv)->num_sprites[(__pipe)] + 1;	\

drivers/gpu/drm/i915/intel_dp_link_training.c

@@ -139,6 +139,11 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp)
 	intel_dp_compute_rate(intel_dp, intel_dp->link_rate,
 			      &link_bw, &rate_select);
 
+	if (link_bw)
+		DRM_DEBUG_KMS("Using LINK_BW_SET value %02x\n", link_bw);
+	else
+		DRM_DEBUG_KMS("Using LINK_RATE_SET value %02x\n", rate_select);
+
 	/* Write the link configuration data */
 	link_config[0] = link_bw;
 	link_config[1] = intel_dp->lane_count;

drivers/gpu/drm/i915/intel_dpll_mgr.h

@@ -103,6 +103,32 @@ enum intel_dpll_id {
 	 * @DPLL_ID_SKL_DPLL3: SKL and later DPLL3
 	 */
 	DPLL_ID_SKL_DPLL3 = 3,
+
+
+	/**
+	 * @DPLL_ID_ICL_DPLL0: ICL combo PHY DPLL0
+	 */
+	DPLL_ID_ICL_DPLL0 = 0,
+	/**
+	 * @DPLL_ID_ICL_DPLL1: ICL combo PHY DPLL1
+	 */
+	DPLL_ID_ICL_DPLL1 = 1,
+	/**
+	 * @DPLL_ID_ICL_MGPLL1: ICL MG PLL 1 port 1 (C)
+	 */
+	DPLL_ID_ICL_MGPLL1 = 2,
+	/**
+	 * @DPLL_ID_ICL_MGPLL2: ICL MG PLL 1 port 2 (D)
+	 */
+	DPLL_ID_ICL_MGPLL2 = 3,
+	/**
+	 * @DPLL_ID_ICL_MGPLL3: ICL MG PLL 1 port 3 (E)
+	 */
+	DPLL_ID_ICL_MGPLL3 = 4,
+	/**
+	 * @DPLL_ID_ICL_MGPLL4: ICL MG PLL 1 port 4 (F)
+	 */
+	DPLL_ID_ICL_MGPLL4 = 5,
 };
 #define I915_NUM_PLLS 6
 
@@ -135,6 +161,21 @@ struct intel_dpll_hw_state {
 	/* bxt */
 	uint32_t ebb0, ebb4, pll0, pll1, pll2, pll3, pll6, pll8, pll9, pll10,
 		 pcsdw12;
+
+	/*
+	 * ICL uses the following, already defined:
+	 * uint32_t cfgcr0, cfgcr1;
+	 */
+	uint32_t mg_refclkin_ctl;
+	uint32_t mg_clktop2_coreclkctl1;
+	uint32_t mg_clktop2_hsclkctl;
+	uint32_t mg_pll_div0;
+	uint32_t mg_pll_div1;
+	uint32_t mg_pll_lf;
+	uint32_t mg_pll_frac_lock;
+	uint32_t mg_pll_ssc;
+	uint32_t mg_pll_bias;
+	uint32_t mg_pll_tdc_coldst_bias;
 };
 
 /**

drivers/gpu/drm/i915/intel_drv.h

@@ -56,6 +56,8 @@
 	for (;;) {							\
 		const bool expired__ = ktime_after(ktime_get_raw(), end__); \
 		OP;							\
+		/* Guarantee COND check prior to timeout */		\
+		barrier();						\
 		if (COND) {						\
 			ret__ = 0;					\
 			break;						\
@@ -96,6 +98,8 @@
 		u64 now = local_clock(); \
 		if (!(ATOMIC)) \
 			preempt_enable(); \
+		/* Guarantee COND check prior to timeout */ \
+		barrier(); \
 		if (COND) { \
 			ret = 0; \
 			break; \
@@ -140,6 +144,10 @@
 #define KHz(x) (1000 * (x))
 #define MHz(x) KHz(1000 * (x))
 
+#define KBps(x) (1000 * (x))
+#define MBps(x) KBps(1000 * (x))
+#define GBps(x) ((u64)1000 * MBps((x)))
+
 /*
  * Display related stuff
  */
@@ -882,6 +890,7 @@ struct intel_crtc_state {
 
 	/* bitmask of visible planes (enum plane_id) */
 	u8 active_planes;
+	u8 nv12_planes;
 
 	/* HDMI scrambling status */
 	bool hdmi_scrambling;
@@ -1329,6 +1338,9 @@ void intel_check_cpu_fifo_underruns(struct drm_i915_private *dev_priv);
 void intel_check_pch_fifo_underruns(struct drm_i915_private *dev_priv);
 
 /* i915_irq.c */
+bool gen11_reset_one_iir(struct drm_i915_private * const i915,
+			 const unsigned int bank,
+			 const unsigned int bit);
 void gen5_enable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen5_disable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void gen6_mask_pm_irq(struct drm_i915_private *dev_priv, u32 mask);
@@ -1398,6 +1410,12 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp);
 u8 intel_ddi_dp_voltage_max(struct intel_encoder *encoder);
 int intel_ddi_toggle_hdcp_signalling(struct intel_encoder *intel_encoder,
 				     bool enable);
+void icl_map_plls_to_ports(struct drm_crtc *crtc,
+			   struct intel_crtc_state *crtc_state,
+			   struct drm_atomic_state *old_state);
+void icl_unmap_plls_to_ports(struct drm_crtc *crtc,
+			     struct intel_crtc_state *crtc_state,
+			     struct drm_atomic_state *old_state);
 
 unsigned int intel_fb_align_height(const struct drm_framebuffer *fb,
 				   int plane, unsigned int height);
@@ -1580,8 +1598,6 @@ void bxt_enable_dc9(struct drm_i915_private *dev_priv);
 void bxt_disable_dc9(struct drm_i915_private *dev_priv);
 void gen9_enable_dc5(struct drm_i915_private *dev_priv);
 unsigned int skl_cdclk_get_vco(unsigned int freq);
-void skl_enable_dc6(struct drm_i915_private *dev_priv);
-void skl_disable_dc6(struct drm_i915_private *dev_priv);
 void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_state *pipe_config);
 void intel_dp_set_m_n(struct intel_crtc *crtc, enum link_m_n_set m_n);
@@ -1901,6 +1917,8 @@ void intel_psr_single_frame_update(struct drm_i915_private *dev_priv,
 				   unsigned frontbuffer_bits);
 void intel_psr_compute_config(struct intel_dp *intel_dp,
 			      struct intel_crtc_state *crtc_state);
+void intel_psr_irq_control(struct drm_i915_private *dev_priv, bool debug);
+void intel_psr_irq_handler(struct drm_i915_private *dev_priv, u32 psr_iir);
 
 /* intel_runtime_pm.c */
 int intel_power_domains_init(struct drm_i915_private *);
@@ -1924,6 +1942,8 @@ bool intel_display_power_get_if_enabled(struct drm_i915_private *dev_priv,
 					enum intel_display_power_domain domain);
 void intel_display_power_put(struct drm_i915_private *dev_priv,
 			     enum intel_display_power_domain domain);
+void icl_dbuf_slices_update(struct drm_i915_private *dev_priv,
+			    u8 req_slices);
 
 static inline void
 assert_rpm_device_not_suspended(struct drm_i915_private *dev_priv)
@@ -2062,6 +2082,8 @@ bool skl_plane_get_hw_state(struct intel_plane *plane);
 bool skl_plane_has_ccs(struct drm_i915_private *dev_priv,
 		       enum pipe pipe, enum plane_id plane_id);
 bool intel_format_is_yuv(uint32_t format);
+bool skl_plane_has_planar(struct drm_i915_private *dev_priv,
+			  enum pipe pipe, enum plane_id plane_id);
 
 /* intel_tv.c */
 void intel_tv_init(struct drm_i915_private *dev_priv);

drivers/gpu/drm/i915/intel_dsi_vbt.c

@@ -647,6 +647,11 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 	/* prepare count */
 	prepare_cnt = DIV_ROUND_UP(ths_prepare_ns * ui_den, ui_num * mul);
 
+	if (prepare_cnt > PREPARE_CNT_MAX) {
+		DRM_DEBUG_KMS("prepare count too high %u\n", prepare_cnt);
+		prepare_cnt = PREPARE_CNT_MAX;
+	}
+
 	/* exit zero count */
 	exit_zero_cnt = DIV_ROUND_UP(
 				(ths_prepare_hszero - ths_prepare_ns) * ui_den,
@@ -662,32 +667,29 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 	if (exit_zero_cnt < (55 * ui_den / ui_num) && (55 * ui_den) % ui_num)
 		exit_zero_cnt += 1;
 
+	if (exit_zero_cnt > EXIT_ZERO_CNT_MAX) {
+		DRM_DEBUG_KMS("exit zero count too high %u\n", exit_zero_cnt);
+		exit_zero_cnt = EXIT_ZERO_CNT_MAX;
+	}
+
 	/* clk zero count */
 	clk_zero_cnt = DIV_ROUND_UP(
 				(tclk_prepare_clkzero -	ths_prepare_ns)
 				* ui_den, ui_num * mul);
 
+	if (clk_zero_cnt > CLK_ZERO_CNT_MAX) {
+		DRM_DEBUG_KMS("clock zero count too high %u\n", clk_zero_cnt);
+		clk_zero_cnt = CLK_ZERO_CNT_MAX;
+	}
+
 	/* trail count */
 	tclk_trail_ns = max(mipi_config->tclk_trail, mipi_config->ths_trail);
 	trail_cnt = DIV_ROUND_UP(tclk_trail_ns * ui_den, ui_num * mul);
 
-	if (prepare_cnt > PREPARE_CNT_MAX ||
-		exit_zero_cnt > EXIT_ZERO_CNT_MAX ||
-		clk_zero_cnt > CLK_ZERO_CNT_MAX ||
-		trail_cnt > TRAIL_CNT_MAX)
-		DRM_DEBUG_DRIVER("Values crossing maximum limits, restricting to max values\n");
-
-	if (prepare_cnt > PREPARE_CNT_MAX)
-		prepare_cnt = PREPARE_CNT_MAX;
-
-	if (exit_zero_cnt > EXIT_ZERO_CNT_MAX)
-		exit_zero_cnt = EXIT_ZERO_CNT_MAX;
-
-	if (clk_zero_cnt > CLK_ZERO_CNT_MAX)
-		clk_zero_cnt = CLK_ZERO_CNT_MAX;
-
-	if (trail_cnt > TRAIL_CNT_MAX)
+	if (trail_cnt > TRAIL_CNT_MAX) {
+		DRM_DEBUG_KMS("trail count too high %u\n", trail_cnt);
 		trail_cnt = TRAIL_CNT_MAX;
+	}
 
 	/* B080 */
 	intel_dsi->dphy_reg = exit_zero_cnt << 24 | trail_cnt << 16 |

drivers/gpu/drm/i915/intel_guc_ads.c

@@ -121,7 +121,8 @@ int intel_guc_ads_create(struct intel_guc *guc)
 	 * to find it. Note that we have to skip our header (1 page),
 	 * because our GuC shared data is there.
 	 */
-	kernel_ctx_vma = dev_priv->kernel_context->engine[RCS].state;
+	kernel_ctx_vma = to_intel_context(dev_priv->kernel_context,
+					  dev_priv->engine[RCS])->state;
 	blob->ads.golden_context_lrca =
 		intel_guc_ggtt_offset(guc, kernel_ctx_vma) + skipped_offset;
 

drivers/gpu/drm/i915/intel_guc_fwif.h

@@ -23,9 +23,6 @@
 #ifndef _INTEL_GUC_FWIF_H
 #define _INTEL_GUC_FWIF_H
 
-#define GUC_CORE_FAMILY_GEN9		12
-#define GUC_CORE_FAMILY_UNKNOWN		0x7fffffff
-
 #define GUC_CLIENT_PRIORITY_KMD_HIGH	0
 #define GUC_CLIENT_PRIORITY_HIGH	1
 #define GUC_CLIENT_PRIORITY_KMD_NORMAL	2
@@ -82,8 +79,6 @@
 #define GUC_CTL_ARAT_LOW		2
 
 #define GUC_CTL_DEVICE_INFO		3
-#define   GUC_CTL_GT_TYPE_SHIFT		0
-#define   GUC_CTL_CORE_FAMILY_SHIFT	7
 
 #define GUC_CTL_LOG_PARAMS		4
 #define   GUC_LOG_VALID			(1 << 0)