Merge tag 'drm-intel-next-2019-03-20' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

UAPI Changes:
- Report an error early instead of SIGBUS later when mmap beyond BO size

Core Changes:
- This includes backmerge of drm-next and two merges of Maarten's
  topic/hdr-formats

Driver Changes:
- Add Comet Lake (Gen9) PCI IDs to Coffee Lake ID list (Anusha)
- Add missing ICL PCI ID (Jose)
- Fix legacy gamma mode for ICL (Ville)
- Assume eDP is present on port A when there is no VBT (Thomas)
- Corrections to eDP training patterns (Jose)
- Fix PSR2 selective update corruption after PSR1 setup (Jose)
- Fix CRC mismatch error for DP link layer compliance (Aditya)
- Fix CNL DPLL readout and clean up code (Ville)
- Turn off the CUS when turning off a HDR plane (Ville)
- Avoid a race with execlist tasklet during race (Chris)
- Add missing CSC readout and clean up code (Ville)
- Avoid unnecessary wakeref during debugfs/drop_caches/set (Chris, Caz)
- Hold references to ring/HW context/context explicitly when used (Chris)

- Assume next platforms inherit old platform (Rodrigo)
- Use HWS indices rather than addresses for breadcrumbs (Chris)
- Add REG_BIT/REG_GENMASK and REG_FIELD_PREP macros (Jani)
- Convert crept in C99 types to kernel fixed size types (Jani)
- Avoid passing full dev_priv in forcewake functions (Daniele)
- Reset GuC on GPU reset (Sujaritha)
- Rework MG and Combo PLLs to vfuncs (Lucas)
- Explicitly track ppGTT size (Chris, Bob)
- Coding style improvements and code modularization (Ville)
- Selftest and debugging improvements (Chris)
Signed-off-by: Dave Airlie <airlied@redhat.com>

# Conflicts:
#	drivers/gpu/drm/i915/intel_hdmi.c
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190325124925.GA12726@jlahtine-desk.ger.corp.intel.com

drivers/gpu/drm/i915/Makefile

@@ -56,6 +56,15 @@ i915-$(CONFIG_COMPAT)   += i915_ioc32.o
 i915-$(CONFIG_DEBUG_FS) += i915_debugfs.o intel_pipe_crc.o
 i915-$(CONFIG_PERF_EVENTS) += i915_pmu.o
 
+# Test the headers are compilable as standalone units
+i915-$(CONFIG_DRM_I915_WERROR) += \
+	test_i915_active_types_standalone.o \
+	test_i915_gem_context_types_standalone.o \
+	test_i915_timeline_types_standalone.o \
+	test_intel_context_types_standalone.o \
+	test_intel_engine_types_standalone.o \
+	test_intel_workarounds_types_standalone.o
+
 # GEM code
 i915-y += \
 	  i915_active.o \
@@ -77,6 +86,7 @@ i915-y += \
 	  i915_gem_tiling.o \
 	  i915_gem_userptr.o \
 	  i915_gemfs.o \
+	  i915_globals.o \
 	  i915_query.o \
 	  i915_request.o \
 	  i915_scheduler.o \
@@ -84,6 +94,7 @@ i915-y += \
 	  i915_trace_points.o \
 	  i915_vma.o \
 	  intel_breadcrumbs.o \
+	  intel_context.o \
 	  intel_engine_cs.o \
 	  intel_hangcheck.o \
 	  intel_lrc.o \

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

@@ -391,12 +391,12 @@ struct cmd_info {
 #define F_POST_HANDLE	(1<<2)
 	u32 flag;
 
-#define R_RCS	(1 << RCS)
-#define R_VCS1  (1 << VCS)
-#define R_VCS2  (1 << VCS2)
+#define R_RCS	BIT(RCS0)
+#define R_VCS1  BIT(VCS0)
+#define R_VCS2  BIT(VCS1)
 #define R_VCS	(R_VCS1 | R_VCS2)
-#define R_BCS	(1 << BCS)
-#define R_VECS	(1 << VECS)
+#define R_BCS	BIT(BCS0)
+#define R_VECS	BIT(VECS0)
 #define R_ALL (R_RCS | R_VCS | R_BCS | R_VECS)
 	/* rings that support this cmd: BLT/RCS/VCS/VECS */
 	u16 rings;
@@ -558,7 +558,7 @@ static const struct decode_info decode_info_vebox = {
 };
 
 static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
-	[RCS] = {
+	[RCS0] = {
 		&decode_info_mi,
 		NULL,
 		NULL,
@@ -569,7 +569,7 @@ static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
 		NULL,
 	},
 
-	[VCS] = {
+	[VCS0] = {
 		&decode_info_mi,
 		NULL,
 		NULL,
@@ -580,7 +580,7 @@ static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
 		NULL,
 	},
 
-	[BCS] = {
+	[BCS0] = {
 		&decode_info_mi,
 		NULL,
 		&decode_info_2d,
@@ -591,7 +591,7 @@ static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
 		NULL,
 	},
 
-	[VECS] = {
+	[VECS0] = {
 		&decode_info_mi,
 		NULL,
 		NULL,
@@ -602,7 +602,7 @@ static const struct decode_info *ring_decode_info[I915_NUM_ENGINES][8] = {
 		NULL,
 	},
 
-	[VCS2] = {
+	[VCS1] = {
 		&decode_info_mi,
 		NULL,
 		NULL,
@@ -631,8 +631,7 @@ static inline const struct cmd_info *find_cmd_entry(struct intel_gvt *gvt,
 	struct cmd_entry *e;
 
 	hash_for_each_possible(gvt->cmd_table, e, hlist, opcode) {
-		if ((opcode == e->info->opcode) &&
-				(e->info->rings & (1 << ring_id)))
+		if (opcode == e->info->opcode && e->info->rings & BIT(ring_id))
 			return e->info;
 	}
 	return NULL;
@@ -943,15 +942,12 @@ static int cmd_handler_lri(struct parser_exec_state *s)
 	struct intel_gvt *gvt = s->vgpu->gvt;
 
 	for (i = 1; i < cmd_len; i += 2) {
-		if (IS_BROADWELL(gvt->dev_priv) &&
-				(s->ring_id != RCS)) {
-			if (s->ring_id == BCS &&
-					cmd_reg(s, i) ==
-					i915_mmio_reg_offset(DERRMR))
+		if (IS_BROADWELL(gvt->dev_priv) && s->ring_id != RCS0) {
+			if (s->ring_id == BCS0 &&
+			    cmd_reg(s, i) == i915_mmio_reg_offset(DERRMR))
 				ret |= 0;
 			else
-				ret |= (cmd_reg_inhibit(s, i)) ?
-					-EBADRQC : 0;
+				ret |= cmd_reg_inhibit(s, i) ? -EBADRQC : 0;
 		}
 		if (ret)
 			break;
@@ -1047,27 +1043,27 @@ struct cmd_interrupt_event {
 };
 
 static struct cmd_interrupt_event cmd_interrupt_events[] = {
-	[RCS] = {
+	[RCS0] = {
 		.pipe_control_notify = RCS_PIPE_CONTROL,
 		.mi_flush_dw = INTEL_GVT_EVENT_RESERVED,
 		.mi_user_interrupt = RCS_MI_USER_INTERRUPT,
 	},
-	[BCS] = {
+	[BCS0] = {
 		.pipe_control_notify = INTEL_GVT_EVENT_RESERVED,
 		.mi_flush_dw = BCS_MI_FLUSH_DW,
 		.mi_user_interrupt = BCS_MI_USER_INTERRUPT,
 	},
-	[VCS] = {
+	[VCS0] = {
 		.pipe_control_notify = INTEL_GVT_EVENT_RESERVED,
 		.mi_flush_dw = VCS_MI_FLUSH_DW,
 		.mi_user_interrupt = VCS_MI_USER_INTERRUPT,
 	},
-	[VCS2] = {
+	[VCS1] = {
 		.pipe_control_notify = INTEL_GVT_EVENT_RESERVED,
 		.mi_flush_dw = VCS2_MI_FLUSH_DW,
 		.mi_user_interrupt = VCS2_MI_USER_INTERRUPT,
 	},
-	[VECS] = {
+	[VECS0] = {
 		.pipe_control_notify = INTEL_GVT_EVENT_RESERVED,
 		.mi_flush_dw = VECS_MI_FLUSH_DW,
 		.mi_user_interrupt = VECS_MI_USER_INTERRUPT,

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

@@ -153,7 +153,7 @@ static struct drm_i915_gem_object *vgpu_create_gem(struct drm_device *dev,
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_object *obj;
 
-	obj = i915_gem_object_alloc(dev_priv);
+	obj = i915_gem_object_alloc();
 	if (obj == NULL)
 		return NULL;
 

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

@@ -47,17 +47,16 @@
 		((a)->lrca == (b)->lrca))
 
 static int context_switch_events[] = {
-	[RCS] = RCS_AS_CONTEXT_SWITCH,
-	[BCS] = BCS_AS_CONTEXT_SWITCH,
-	[VCS] = VCS_AS_CONTEXT_SWITCH,
-	[VCS2] = VCS2_AS_CONTEXT_SWITCH,
-	[VECS] = VECS_AS_CONTEXT_SWITCH,
+	[RCS0]  = RCS_AS_CONTEXT_SWITCH,
+	[BCS0]  = BCS_AS_CONTEXT_SWITCH,
+	[VCS0]  = VCS_AS_CONTEXT_SWITCH,
+	[VCS1]  = VCS2_AS_CONTEXT_SWITCH,
+	[VECS0] = VECS_AS_CONTEXT_SWITCH,
 };
 
-static int ring_id_to_context_switch_event(int ring_id)
+static int ring_id_to_context_switch_event(unsigned int ring_id)
 {
-	if (WARN_ON(ring_id < RCS ||
-		    ring_id >= ARRAY_SIZE(context_switch_events)))
+	if (WARN_ON(ring_id >= ARRAY_SIZE(context_switch_events)))
 		return -EINVAL;
 
 	return context_switch_events[ring_id];
@@ -411,7 +410,7 @@ static int complete_execlist_workload(struct intel_vgpu_workload *workload)
 	gvt_dbg_el("complete workload %p status %d\n", workload,
 			workload->status);
 
-	if (workload->status || (vgpu->resetting_eng & ENGINE_MASK(ring_id)))
+	if (workload->status || (vgpu->resetting_eng & BIT(ring_id)))
 		goto out;
 
 	if (!list_empty(workload_q_head(vgpu, ring_id))) {

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

@@ -323,25 +323,25 @@ static int gdrst_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 	} else {
 		if (data & GEN6_GRDOM_RENDER) {
 			gvt_dbg_mmio("vgpu%d: request RCS reset\n", vgpu->id);
-			engine_mask |= (1 << RCS);
+			engine_mask |= BIT(RCS0);
 		}
 		if (data & GEN6_GRDOM_MEDIA) {
 			gvt_dbg_mmio("vgpu%d: request VCS reset\n", vgpu->id);
-			engine_mask |= (1 << VCS);
+			engine_mask |= BIT(VCS0);
 		}
 		if (data & GEN6_GRDOM_BLT) {
 			gvt_dbg_mmio("vgpu%d: request BCS Reset\n", vgpu->id);
-			engine_mask |= (1 << BCS);
+			engine_mask |= BIT(BCS0);
 		}
 		if (data & GEN6_GRDOM_VECS) {
 			gvt_dbg_mmio("vgpu%d: request VECS Reset\n", vgpu->id);
-			engine_mask |= (1 << VECS);
+			engine_mask |= BIT(VECS0);
 		}
 		if (data & GEN8_GRDOM_MEDIA2) {
 			gvt_dbg_mmio("vgpu%d: request VCS2 Reset\n", vgpu->id);
-			if (HAS_BSD2(vgpu->gvt->dev_priv))
-				engine_mask |= (1 << VCS2);
+			engine_mask |= BIT(VCS1);
 		}
+		engine_mask &= INTEL_INFO(vgpu->gvt->dev_priv)->engine_mask;
 	}
 
 	/* vgpu_lock already hold by emulate mmio r/w */
@@ -1704,7 +1704,7 @@ static int ring_mode_mmio_write(struct intel_vgpu *vgpu, unsigned int offset,
 			return 0;
 
 		ret = intel_vgpu_select_submission_ops(vgpu,
-			       ENGINE_MASK(ring_id),
+			       BIT(ring_id),
 			       INTEL_VGPU_EXECLIST_SUBMISSION);
 		if (ret)
 			return ret;
@@ -1724,19 +1724,19 @@ static int gvt_reg_tlb_control_handler(struct intel_vgpu *vgpu,
 
 	switch (offset) {
 	case 0x4260:
-		id = RCS;
+		id = RCS0;
 		break;
 	case 0x4264:
-		id = VCS;
+		id = VCS0;
 		break;
 	case 0x4268:
-		id = VCS2;
+		id = VCS1;
 		break;
 	case 0x426c:
-		id = BCS;
+		id = BCS0;
 		break;
 	case 0x4270:
-		id = VECS;
+		id = VECS0;
 		break;
 	default:
 		return -EINVAL;
@@ -1793,7 +1793,7 @@ static int ring_reset_ctl_write(struct intel_vgpu *vgpu,
 	MMIO_F(prefix(BLT_RING_BASE), s, f, am, rm, d, r, w); \
 	MMIO_F(prefix(GEN6_BSD_RING_BASE), s, f, am, rm, d, r, w); \
 	MMIO_F(prefix(VEBOX_RING_BASE), s, f, am, rm, d, r, w); \
-	if (HAS_BSD2(dev_priv)) \
+	if (HAS_ENGINE(dev_priv, VCS1)) \
 		MMIO_F(prefix(GEN8_BSD2_RING_BASE), s, f, am, rm, d, r, w); \
 } while (0)
 

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

@@ -536,7 +536,7 @@ static void gen8_init_irq(
 	SET_BIT_INFO(irq, 4, VCS_MI_FLUSH_DW, INTEL_GVT_IRQ_INFO_GT1);
 	SET_BIT_INFO(irq, 8, VCS_AS_CONTEXT_SWITCH, INTEL_GVT_IRQ_INFO_GT1);
 
-	if (HAS_BSD2(gvt->dev_priv)) {
+	if (HAS_ENGINE(gvt->dev_priv, VCS1)) {
 		SET_BIT_INFO(irq, 16, VCS2_MI_USER_INTERRUPT,
 			INTEL_GVT_IRQ_INFO_GT1);
 		SET_BIT_INFO(irq, 20, VCS2_MI_FLUSH_DW,

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

@@ -93,7 +93,7 @@ static void sr_oa_regs(struct intel_vgpu_workload *workload,
 		i915_mmio_reg_offset(EU_PERF_CNTL6),
 	};
 
-	if (workload->ring_id != RCS)
+	if (workload->ring_id != RCS0)
 		return;
 
 	if (save) {
@@ -149,7 +149,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	COPY_REG_MASKED(ctx_ctrl);
 	COPY_REG(ctx_timestamp);
 
-	if (ring_id == RCS) {
+	if (ring_id == RCS0) {
 		COPY_REG(bb_per_ctx_ptr);
 		COPY_REG(rcs_indirect_ctx);
 		COPY_REG(rcs_indirect_ctx_offset);
@@ -177,7 +177,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
-	if (IS_BROADWELL(gvt->dev_priv) && ring_id == RCS)
+	if (IS_BROADWELL(gvt->dev_priv) && ring_id == RCS0)
 		context_page_num = 19;
 
 	i = 2;
@@ -440,8 +440,7 @@ int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload)
 	if (ret)
 		goto err_unpin;
 
-	if ((workload->ring_id == RCS) &&
-	    (workload->wa_ctx.indirect_ctx.size != 0)) {
+	if (workload->ring_id == RCS0 && workload->wa_ctx.indirect_ctx.size) {
 		ret = intel_gvt_scan_and_shadow_wa_ctx(&workload->wa_ctx);
 		if (ret)
 			goto err_shadow;
@@ -791,7 +790,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 	context_page_num = rq->engine->context_size;
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
-	if (IS_BROADWELL(gvt->dev_priv) && rq->engine->id == RCS)
+	if (IS_BROADWELL(gvt->dev_priv) && rq->engine->id == RCS0)
 		context_page_num = 19;
 
 	i = 2;
@@ -891,8 +890,8 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 				workload->status = 0;
 		}
 
-		if (!workload->status && !(vgpu->resetting_eng &
-					   ENGINE_MASK(ring_id))) {
+		if (!workload->status &&
+		    !(vgpu->resetting_eng & BIT(ring_id))) {
 			update_guest_context(workload);
 
 			for_each_set_bit(event, workload->pending_events,
@@ -915,7 +914,7 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 
 	list_del_init(&workload->list);
 
-	if (workload->status || (vgpu->resetting_eng & ENGINE_MASK(ring_id))) {
+	if (workload->status || vgpu->resetting_eng & BIT(ring_id)) {
 		/* if workload->status is not successful means HW GPU
 		 * has occurred GPU hang or something wrong with i915/GVT,
 		 * and GVT won't inject context switch interrupt to guest.
@@ -929,7 +928,7 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id)
 		 * cleaned up during the resetting process later, so doing
 		 * the workload clean up here doesn't have any impact.
 		 **/
-		intel_vgpu_clean_workloads(vgpu, ENGINE_MASK(ring_id));
+		intel_vgpu_clean_workloads(vgpu, BIT(ring_id));
 	}
 
 	workload->complete(workload);
@@ -1102,9 +1101,9 @@ i915_context_ppgtt_root_restore(struct intel_vgpu_submission *s)
 	struct i915_hw_ppgtt *i915_ppgtt = s->shadow_ctx->ppgtt;
 	int i;
 
-	if (i915_vm_is_48bit(&i915_ppgtt->vm))
+	if (i915_vm_is_4lvl(&i915_ppgtt->vm)) {
 		px_dma(&i915_ppgtt->pml4) = s->i915_context_pml4;
-	else {
+	} else {
 		for (i = 0; i < GEN8_3LVL_PDPES; i++)
 			px_dma(i915_ppgtt->pdp.page_directory[i]) =
 						s->i915_context_pdps[i];
@@ -1155,7 +1154,7 @@ i915_context_ppgtt_root_save(struct intel_vgpu_submission *s)
 	struct i915_hw_ppgtt *i915_ppgtt = s->shadow_ctx->ppgtt;
 	int i;
 
-	if (i915_vm_is_48bit(&i915_ppgtt->vm))
+	if (i915_vm_is_4lvl(&i915_ppgtt->vm))
 		s->i915_context_pml4 = px_dma(&i915_ppgtt->pml4);
 	else {
 		for (i = 0; i < GEN8_3LVL_PDPES; i++)
@@ -1438,7 +1437,7 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 	workload->rb_start = start;
 	workload->rb_ctl = ctl;
 
-	if (ring_id == RCS) {
+	if (ring_id == RCS0) {
 		intel_gvt_hypervisor_read_gpa(vgpu, ring_context_gpa +
 			RING_CTX_OFF(bb_per_ctx_ptr.val), &per_ctx, 4);
 		intel_gvt_hypervisor_read_gpa(vgpu, ring_context_gpa +

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

@@ -44,7 +44,7 @@ void populate_pvinfo_page(struct intel_vgpu *vgpu)
 	vgpu_vreg_t(vgpu, vgtif_reg(display_ready)) = 0;
 	vgpu_vreg_t(vgpu, vgtif_reg(vgt_id)) = vgpu->id;
 
-	vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) = VGT_CAPS_FULL_48BIT_PPGTT;
+	vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) = VGT_CAPS_FULL_PPGTT;
 	vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) |= VGT_CAPS_HWSP_EMULATION;
 	vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) |= VGT_CAPS_HUGE_GTT;
 

drivers/gpu/drm/i915/i915_active.c

@@ -6,6 +6,7 @@
 
 #include "i915_drv.h"
 #include "i915_active.h"
+#include "i915_globals.h"
 
 #define BKL(ref) (&(ref)->i915->drm.struct_mutex)
 
@@ -17,6 +18,7 @@
  * nodes from a local slab cache to hopefully reduce the fragmentation.
  */
 static struct i915_global_active {
+	struct i915_global base;
 	struct kmem_cache *slab_cache;
 } global;
 
@@ -285,16 +287,27 @@ void i915_active_retire_noop(struct i915_active_request *active,
 #include "selftests/i915_active.c"
 #endif
 
+static void i915_global_active_shrink(void)
+{
+	kmem_cache_shrink(global.slab_cache);
+}
+
+static void i915_global_active_exit(void)
+{
+	kmem_cache_destroy(global.slab_cache);
+}
+
+static struct i915_global_active global = { {
+	.shrink = i915_global_active_shrink,
+	.exit = i915_global_active_exit,
+} };
+
 int __init i915_global_active_init(void)
 {
 	global.slab_cache = KMEM_CACHE(active_node, SLAB_HWCACHE_ALIGN);
 	if (!global.slab_cache)
 		return -ENOMEM;
 
+	i915_global_register(&global.base);
 	return 0;
 }
-
-void __exit i915_global_active_exit(void)
-{
-	kmem_cache_destroy(global.slab_cache);
-}

drivers/gpu/drm/i915/i915_active.h

@@ -108,19 +108,6 @@ i915_active_request_set_retire_fn(struct i915_active_request *active,
 	active->retire = fn ?: i915_active_retire_noop;
 }
 
-static inline struct i915_request *
-__i915_active_request_peek(const struct i915_active_request *active)
-{
-	/*
-	 * Inside the error capture (running with the driver in an unknown
-	 * state), we want to bend the rules slightly (a lot).
-	 *
-	 * Work is in progress to make it safer, in the meantime this keeps
-	 * the known issue from spamming the logs.
-	 */
-	return rcu_dereference_protected(active->request, 1);
-}
-
 /**
  * i915_active_request_raw - return the active request
  * @active - the active tracker
@@ -419,7 +406,4 @@ void i915_active_fini(struct i915_active *ref);
 static inline void i915_active_fini(struct i915_active *ref) { }
 #endif
 
-int i915_global_active_init(void);
-void i915_global_active_exit(void);
-
 #endif /* _I915_ACTIVE_H_ */

drivers/gpu/drm/i915/i915_cmd_parser.c

@@ -868,8 +868,8 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 	if (!IS_GEN(engine->i915, 7))
 		return;
 
-	switch (engine->id) {
-	case RCS:
+	switch (engine->class) {
+	case RENDER_CLASS:
 		if (IS_HASWELL(engine->i915)) {
 			cmd_tables = hsw_render_ring_cmds;
 			cmd_table_count =
@@ -889,12 +889,12 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 
 		engine->get_cmd_length_mask = gen7_render_get_cmd_length_mask;
 		break;
-	case VCS:
+	case VIDEO_DECODE_CLASS:
 		cmd_tables = gen7_video_cmds;
 		cmd_table_count = ARRAY_SIZE(gen7_video_cmds);
 		engine->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
 		break;
-	case BCS:
+	case COPY_ENGINE_CLASS:
 		if (IS_HASWELL(engine->i915)) {
 			cmd_tables = hsw_blt_ring_cmds;
 			cmd_table_count = ARRAY_SIZE(hsw_blt_ring_cmds);
@@ -913,14 +913,14 @@ void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 
 		engine->get_cmd_length_mask = gen7_blt_get_cmd_length_mask;
 		break;
-	case VECS:
+	case VIDEO_ENHANCEMENT_CLASS:
 		cmd_tables = hsw_vebox_cmds;
 		cmd_table_count = ARRAY_SIZE(hsw_vebox_cmds);
 		/* VECS can use the same length_mask function as VCS */
 		engine->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
 		break;
 	default:
-		MISSING_CASE(engine->id);
+		MISSING_CASE(engine->class);
 		return;
 	}
 

drivers/gpu/drm/i915/i915_debugfs.c

@@ -388,12 +388,9 @@ static void print_context_stats(struct seq_file *m,
 	struct i915_gem_context *ctx;
 
 	list_for_each_entry(ctx, &i915->contexts.list, link) {
-		struct intel_engine_cs *engine;
-		enum intel_engine_id id;
-
-		for_each_engine(engine, i915, id) {
-			struct intel_context *ce = to_intel_context(ctx, engine);
+		struct intel_context *ce;
 
+		list_for_each_entry(ce, &ctx->active_engines, active_link) {
 			if (ce->state)
 				per_file_stats(0, ce->state->obj, &kstats);
 			if (ce->ring)
@@ -1281,14 +1278,11 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 	intel_wakeref_t wakeref;
 	enum intel_engine_id id;
 
+	seq_printf(m, "Reset flags: %lx\n", dev_priv->gpu_error.flags);
 	if (test_bit(I915_WEDGED, &dev_priv->gpu_error.flags))
-		seq_puts(m, "Wedged\n");
+		seq_puts(m, "\tWedged\n");
 	if (test_bit(I915_RESET_BACKOFF, &dev_priv->gpu_error.flags))
-		seq_puts(m, "Reset in progress: struct_mutex backoff\n");
-	if (waitqueue_active(&dev_priv->gpu_error.wait_queue))
-		seq_puts(m, "Waiter holding struct mutex\n");
-	if (waitqueue_active(&dev_priv->gpu_error.reset_queue))
-		seq_puts(m, "struct_mutex blocked for reset\n");
+		seq_puts(m, "\tDevice (global) reset in progress\n");
 
 	if (!i915_modparams.enable_hangcheck) {
 		seq_puts(m, "Hangcheck disabled\n");
@@ -1298,10 +1292,10 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 	with_intel_runtime_pm(dev_priv, wakeref) {
 		for_each_engine(engine, dev_priv, id) {
 			acthd[id] = intel_engine_get_active_head(engine);
-			seqno[id] = intel_engine_get_seqno(engine);
+			seqno[id] = intel_engine_get_hangcheck_seqno(engine);
 		}
 
-		intel_engine_get_instdone(dev_priv->engine[RCS], &instdone);
+		intel_engine_get_instdone(dev_priv->engine[RCS0], &instdone);
 	}
 
 	if (timer_pending(&dev_priv->gpu_error.hangcheck_work.timer))
@@ -1318,8 +1312,9 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 	for_each_engine(engine, dev_priv, id) {
 		seq_printf(m, "%s:\n", engine->name);
 		seq_printf(m, "\tseqno = %x [current %x, last %x], %dms ago\n",
-			   engine->hangcheck.seqno, seqno[id],
-			   intel_engine_last_submit(engine),
+			   engine->hangcheck.last_seqno,
+			   seqno[id],
+			   engine->hangcheck.next_seqno,
 			   jiffies_to_msecs(jiffies -
 					    engine->hangcheck.action_timestamp));
 
@@ -1327,7 +1322,7 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 			   (long long)engine->hangcheck.acthd,
 			   (long long)acthd[id]);
 
-		if (engine->id == RCS) {
+		if (engine->id == RCS0) {
 			seq_puts(m, "\tinstdone read =\n");
 
 			i915_instdone_info(dev_priv, m, &instdone);
@@ -1882,9 +1877,7 @@ static int i915_context_status(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	struct drm_device *dev = &dev_priv->drm;
-	struct intel_engine_cs *engine;
 	struct i915_gem_context *ctx;
-	enum intel_engine_id id;
 	int ret;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
@@ -1892,6 +1885,8 @@ static int i915_context_status(struct seq_file *m, void *unused)
 		return ret;
 
 	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
+		struct intel_context *ce;
+
 		seq_puts(m, "HW context ");
 		if (!list_empty(&ctx->hw_id_link))
 			seq_printf(m, "%x [pin %u]", ctx->hw_id,
@@ -1914,11 +1909,8 @@ static int i915_context_status(struct seq_file *m, void *unused)
 		seq_putc(m, ctx->remap_slice ? 'R' : 'r');
 		seq_putc(m, '\n');
 
-		for_each_engine(engine, dev_priv, id) {
-			struct intel_context *ce =
-				to_intel_context(ctx, engine);
-
-			seq_printf(m, "%s: ", engine->name);
+		list_for_each_entry(ce, &ctx->active_engines, active_link) {
+			seq_printf(m, "%s: ", ce->engine->name);
 			if (ce->state)
 				describe_obj(m, ce->state->obj);
 			if (ce->ring)
@@ -2023,11 +2015,9 @@ static const char *rps_power_to_str(unsigned int power)
 static int i915_rps_boost_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
 	struct intel_rps *rps = &dev_priv->gt_pm.rps;
 	u32 act_freq = rps->cur_freq;
 	intel_wakeref_t wakeref;
-	struct drm_file *file;
 
 	with_intel_runtime_pm_if_in_use(dev_priv, wakeref) {
 		if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
@@ -2061,22 +2051,7 @@ static int i915_rps_boost_info(struct seq_file *m, void *data)
 		   intel_gpu_freq(dev_priv, rps->efficient_freq),
 		   intel_gpu_freq(dev_priv, rps->boost_freq));
 
-	mutex_lock(&dev->filelist_mutex);
-	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
-		struct drm_i915_file_private *file_priv = file->driver_priv;
-		struct task_struct *task;
-
-		rcu_read_lock();
-		task = pid_task(file->pid, PIDTYPE_PID);
-		seq_printf(m, "%s [%d]: %d boosts\n",
-			   task ? task->comm : "<unknown>",
-			   task ? task->pid : -1,
-			   atomic_read(&file_priv->rps_client.boosts));
-		rcu_read_unlock();
-	}
-	seq_printf(m, "Kernel (anonymous) boosts: %d\n",
-		   atomic_read(&rps->boosts));
-	mutex_unlock(&dev->filelist_mutex);
+	seq_printf(m, "Wait boosts: %d\n", atomic_read(&rps->boosts));
 
 	if (INTEL_GEN(dev_priv) >= 6 &&
 	    rps->enabled &&
@@ -2607,7 +2582,6 @@ static int
 i915_edp_psr_debug_set(void *data, u64 val)
 {
 	struct drm_i915_private *dev_priv = data;
-	struct drm_modeset_acquire_ctx ctx;
 	intel_wakeref_t wakeref;
 	int ret;
 
@@ -2618,18 +2592,7 @@ i915_edp_psr_debug_set(void *data, u64 val)
 
 	wakeref = intel_runtime_pm_get(dev_priv);
 
-	drm_modeset_acquire_init(&ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE);
-
-retry:
-	ret = intel_psr_set_debugfs_mode(dev_priv, &ctx, val);
-	if (ret == -EDEADLK) {
-		ret = drm_modeset_backoff(&ctx);
-		if (!ret)
-			goto retry;
-	}
-
-	drm_modeset_drop_locks(&ctx);
-	drm_modeset_acquire_fini(&ctx);
+	ret = intel_psr_debug_set(dev_priv, val);
 
 	intel_runtime_pm_put(dev_priv, wakeref);
 
@@ -2686,8 +2649,7 @@ static int i915_runtime_pm_status(struct seq_file *m, void *unused)
 	seq_printf(m, "Runtime power status: %s\n",
 		   enableddisabled(!dev_priv->power_domains.wakeref));
 
-	seq_printf(m, "GPU idle: %s (epoch %u)\n",
-		   yesno(!dev_priv->gt.awake), dev_priv->gt.epoch);
+	seq_printf(m, "GPU idle: %s\n", yesno(!dev_priv->gt.awake));
 	seq_printf(m, "IRQs disabled: %s\n",
 		   yesno(!intel_irqs_enabled(dev_priv)));
 #ifdef CONFIG_PM
@@ -3123,8 +3085,7 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 
 	wakeref = intel_runtime_pm_get(dev_priv);
 
-	seq_printf(m, "GT awake? %s (epoch %u)\n",
-		   yesno(dev_priv->gt.awake), dev_priv->gt.epoch);
+	seq_printf(m, "GT awake? %s\n", yesno(dev_priv->gt.awake));
 	seq_printf(m, "Global active requests: %d\n",
 		   dev_priv->gt.active_requests);
 	seq_printf(m, "CS timestamp frequency: %u kHz\n",
@@ -3211,7 +3172,7 @@ static int i915_shared_dplls_info(struct seq_file *m, void *unused)
 static int i915_wa_registers(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *i915 = node_to_i915(m->private);
-	const struct i915_wa_list *wal = &i915->engine[RCS]->ctx_wa_list;
+	const struct i915_wa_list *wal = &i915->engine[RCS0]->ctx_wa_list;
 	struct i915_wa *wa;
 	unsigned int i;
 
@@ -3865,11 +3826,18 @@ static const struct file_operations i915_cur_wm_latency_fops = {
 static int
 i915_wedged_get(void *data, u64 *val)
 {
-	struct drm_i915_private *dev_priv = data;
+	int ret = i915_terminally_wedged(data);
 
-	*val = i915_terminally_wedged(&dev_priv->gpu_error);
-
-	return 0;
+	switch (ret) {
+	case -EIO:
+		*val = 1;
+		return 0;
+	case 0:
+		*val = 0;
+		return 0;
+	default:
+		return ret;
+	}
 }
 
 static int
@@ -3877,16 +3845,9 @@ i915_wedged_set(void *data, u64 val)
 {
 	struct drm_i915_private *i915 = data;
 
-	/*
-	 * There is no safeguard against this debugfs entry colliding
-	 * with the hangcheck calling same i915_handle_error() in
-	 * parallel, causing an explosion. For now we assume that the
-	 * test harness is responsible enough not to inject gpu hangs
-	 * while it is writing to 'i915_wedged'
-	 */
-
-	if (i915_reset_backoff(&i915->gpu_error))
-		return -EAGAIN;
+	/* Flush any previous reset before applying for a new one */
+	wait_event(i915->gpu_error.reset_queue,
+		   !test_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags));
 
 	i915_handle_error(i915, val, I915_ERROR_CAPTURE,
 			  "Manually set wedged engine mask = %llx", val);
@@ -3927,12 +3888,9 @@ static int
 i915_drop_caches_set(void *data, u64 val)
 {
 	struct drm_i915_private *i915 = data;
-	intel_wakeref_t wakeref;
-	int ret = 0;
 
 	DRM_DEBUG("Dropping caches: 0x%08llx [0x%08llx]\n",
 		  val, val & DROP_ALL);
-	wakeref = intel_runtime_pm_get(i915);
 
 	if (val & DROP_RESET_ACTIVE &&
 	    wait_for(intel_engines_are_idle(i915), I915_IDLE_ENGINES_TIMEOUT))
@@ -3941,9 +3899,11 @@ i915_drop_caches_set(void *data, u64 val)
 	/* No need to check and wait for gpu resets, only libdrm auto-restarts
 	 * on ioctls on -EAGAIN. */
 	if (val & (DROP_ACTIVE | DROP_RETIRE | DROP_RESET_SEQNO)) {
+		int ret;
+
 		ret = mutex_lock_interruptible(&i915->drm.struct_mutex);
 		if (ret)
-			goto out;
+			return ret;
 
 		if (val & DROP_ACTIVE)
 			ret = i915_gem_wait_for_idle(i915,
@@ -3957,7 +3917,7 @@ i915_drop_caches_set(void *data, u64 val)
 		mutex_unlock(&i915->drm.struct_mutex);
 	}
 
-	if (val & DROP_RESET_ACTIVE && i915_terminally_wedged(&i915->gpu_error))
+	if (val & DROP_RESET_ACTIVE && i915_terminally_wedged(i915))
 		i915_handle_error(i915, ALL_ENGINES, 0, NULL);
 
 	fs_reclaim_acquire(GFP_KERNEL);
@@ -3982,10 +3942,7 @@ i915_drop_caches_set(void *data, u64 val)
 	if (val & DROP_FREED)
 		i915_gem_drain_freed_objects(i915);
 
-out:
-	intel_runtime_pm_put(i915, wakeref);
-
-	return ret;
+	return 0;
 }
 
 DEFINE_SIMPLE_ATTRIBUTE(i915_drop_caches_fops,

drivers/gpu/drm/i915/i915_gem.h

@@ -75,12 +75,14 @@ struct drm_i915_private;
 
 #define I915_NUM_ENGINES 8
 
+#define I915_GEM_IDLE_TIMEOUT (HZ / 5)
+
 void i915_gem_park(struct drm_i915_private *i915);
 void i915_gem_unpark(struct drm_i915_private *i915);
 
 static inline void __tasklet_disable_sync_once(struct tasklet_struct *t)
 {
-	if (atomic_inc_return(&t->count) == 1)
+	if (!atomic_fetch_inc(&t->count))
 		tasklet_unlock_wait(t);
 }
 
@@ -89,4 +91,9 @@ static inline bool __tasklet_is_enabled(const struct tasklet_struct *t)
 	return !atomic_read(&t->count);
 }
 
+static inline bool __tasklet_enable(struct tasklet_struct *t)
+{
+	return atomic_dec_and_test(&t->count);
+}
+
 #endif /* __I915_GEM_H__ */

drivers/gpu/drm/i915/i915_gem_context.h

@@ -25,210 +25,16 @@
 #ifndef __I915_GEM_CONTEXT_H__
 #define __I915_GEM_CONTEXT_H__
 
-#include <linux/bitops.h>
-#include <linux/list.h>
-#include <linux/radix-tree.h>
+#include "i915_gem_context_types.h"
 
 #include "i915_gem.h"
 #include "i915_scheduler.h"
+#include "intel_context.h"
 #include "intel_device_info.h"
 
-struct pid;
-
 struct drm_device;
 struct drm_file;
 
-struct drm_i915_private;
-struct drm_i915_file_private;
-struct i915_hw_ppgtt;
-struct i915_request;
-struct i915_vma;
-struct intel_ring;
-
-#define DEFAULT_CONTEXT_HANDLE 0
-
-struct intel_context;
-
-struct intel_context_ops {
-	void (*unpin)(struct intel_context *ce);
-	void (*destroy)(struct intel_context *ce);
-};
-
-/*
- * Powergating configuration for a particular (context,engine).
- */
-struct intel_sseu {
-	u8 slice_mask;
-	u8 subslice_mask;
-	u8 min_eus_per_subslice;
-	u8 max_eus_per_subslice;
-};
-
-/**
- * struct i915_gem_context - client state
- *
- * The struct i915_gem_context represents the combined view of the driver and
- * logical hardware state for a particular client.
- */
-struct i915_gem_context {
-	/** i915: i915 device backpointer */
-	struct drm_i915_private *i915;
-
-	/** file_priv: owning file descriptor */
-	struct drm_i915_file_private *file_priv;
-
-	/**
-	 * @ppgtt: unique address space (GTT)
-	 *
-	 * In full-ppgtt mode, each context has its own address space ensuring
-	 * complete seperation of one client from all others.
-	 *
-	 * In other modes, this is a NULL pointer with the expectation that
-	 * the caller uses the shared global GTT.
-	 */
-	struct i915_hw_ppgtt *ppgtt;
-
-	/**
-	 * @pid: process id of creator
-	 *
-	 * Note that who created the context may not be the principle user,
-	 * as the context may be shared across a local socket. However,
-	 * that should only affect the default context, all contexts created
-	 * explicitly by the client are expected to be isolated.
-	 */
-	struct pid *pid;
-
-	/**
-	 * @name: arbitrary name
-	 *
-	 * A name is constructed for the context from the creator's process
-	 * name, pid and user handle in order to uniquely identify the
-	 * context in messages.
-	 */
-	const char *name;
-
-	/** link: place with &drm_i915_private.context_list */
-	struct list_head link;
-	struct llist_node free_link;
-
-	/**
-	 * @ref: reference count
-	 *
-	 * A reference to a context is held by both the client who created it
-	 * and on each request submitted to the hardware using the request
-	 * (to ensure the hardware has access to the state until it has
-	 * finished all pending writes). See i915_gem_context_get() and
-	 * i915_gem_context_put() for access.
-	 */
-	struct kref ref;
-
-	/**
-	 * @rcu: rcu_head for deferred freeing.
-	 */
-	struct rcu_head rcu;
-
-	/**
-	 * @user_flags: small set of booleans controlled by the user
-	 */
-	unsigned long user_flags;
-#define UCONTEXT_NO_ZEROMAP		0
-#define UCONTEXT_NO_ERROR_CAPTURE	1
-#define UCONTEXT_BANNABLE		2
-
-	/**
-	 * @flags: small set of booleans
-	 */
-	unsigned long flags;
-#define CONTEXT_BANNED			0
-#define CONTEXT_CLOSED			1
-#define CONTEXT_FORCE_SINGLE_SUBMISSION	2
-
-	/**
-	 * @hw_id: - unique identifier for the context
-	 *
-	 * The hardware needs to uniquely identify the context for a few
-	 * functions like fault reporting, PASID, scheduling. The
-	 * &drm_i915_private.context_hw_ida is used to assign a unqiue
-	 * id for the lifetime of the context.
-	 *
-	 * @hw_id_pin_count: - number of times this context had been pinned
-	 * for use (should be, at most, once per engine).
-	 *
-	 * @hw_id_link: - all contexts with an assigned id are tracked
-	 * for possible repossession.
-	 */
-	unsigned int hw_id;
-	atomic_t hw_id_pin_count;
-	struct list_head hw_id_link;
-
-	/**
-	 * @user_handle: userspace identifier
-	 *
-	 * A unique per-file identifier is generated from
-	 * &drm_i915_file_private.contexts.
-	 */
-	u32 user_handle;
-
-	struct i915_sched_attr sched;
-
-	/** engine: per-engine logical HW state */
-	struct intel_context {
-		struct i915_gem_context *gem_context;
-		struct intel_engine_cs *active;
-		struct list_head signal_link;
-		struct list_head signals;
-		struct i915_vma *state;
-		struct intel_ring *ring;
-		u32 *lrc_reg_state;
-		u64 lrc_desc;
-		int pin_count;
-
-		/**
-		 * active_tracker: Active tracker for the external rq activity
-		 * on this intel_context object.
-		 */
-		struct i915_active_request active_tracker;
-
-		const struct intel_context_ops *ops;
-
-		/** sseu: Control eu/slice partitioning */
-		struct intel_sseu sseu;
-	} __engine[I915_NUM_ENGINES];
-
-	/** ring_size: size for allocating the per-engine ring buffer */
-	u32 ring_size;
-	/** desc_template: invariant fields for the HW context descriptor */
-	u32 desc_template;
-
-	/** guilty_count: How many times this context has caused a GPU hang. */
-	atomic_t guilty_count;
-	/**
-	 * @active_count: How many times this context was active during a GPU
-	 * hang, but did not cause it.
-	 */
-	atomic_t active_count;
-
-#define CONTEXT_SCORE_GUILTY		10
-#define CONTEXT_SCORE_BAN_THRESHOLD	40
-	/** ban_score: Accumulated score of all hangs caused by this context. */
-	atomic_t ban_score;
-
-	/** remap_slice: Bitmask of cache lines that need remapping */
-	u8 remap_slice;
-
-	/** handles_vma: rbtree to look up our context specific obj/vma for
-	 * the user handle. (user handles are per fd, but the binding is
-	 * per vm, which may be one per context or shared with the global GTT)
-	 */
-	struct radix_tree_root handles_vma;
-
-	/** handles_list: reverse list of all the rbtree entries in use for
-	 * this context, which allows us to free all the allocations on
-	 * context close.
-	 */
-	struct list_head handles_list;
-};
-
 static inline bool i915_gem_context_is_closed(const struct i915_gem_context *ctx)
 {
 	return test_bit(CONTEXT_CLOSED, &ctx->flags);
@@ -270,6 +76,21 @@ static inline void i915_gem_context_clear_bannable(struct i915_gem_context *ctx)
 	clear_bit(UCONTEXT_BANNABLE, &ctx->user_flags);
 }
 
+static inline bool i915_gem_context_is_recoverable(const struct i915_gem_context *ctx)
+{
+	return test_bit(UCONTEXT_RECOVERABLE, &ctx->user_flags);
+}
+
+static inline void i915_gem_context_set_recoverable(struct i915_gem_context *ctx)
+{
+	set_bit(UCONTEXT_RECOVERABLE, &ctx->user_flags);
+}
+
+static inline void i915_gem_context_clear_recoverable(struct i915_gem_context *ctx)
+{
+	clear_bit(UCONTEXT_RECOVERABLE, &ctx->user_flags);
+}
+
 static inline bool i915_gem_context_is_banned(const struct i915_gem_context *ctx)
 {
 	return test_bit(CONTEXT_BANNED, &ctx->flags);
@@ -315,35 +136,6 @@ 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_context *
-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 intel_context *ce)
-{
-	GEM_BUG_ON(!ce->pin_count);
-	ce->pin_count++;
-}
-
-static inline void intel_context_unpin(struct intel_context *ce)
-{
-	GEM_BUG_ON(!ce->pin_count);
-	if (--ce->pin_count)
-		return;
-
-	GEM_BUG_ON(!ce->ops);
-	ce->ops->unpin(ce);
-}
-
 /* 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);
@@ -354,7 +146,8 @@ int i915_gem_context_open(struct drm_i915_private *i915,
 void i915_gem_context_close(struct drm_file *file);
 
 int i915_switch_context(struct i915_request *rq);
-int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv);
+int i915_gem_switch_to_kernel_context(struct drm_i915_private *i915,
+				      unsigned long engine_mask);
 
 void i915_gem_context_release(struct kref *ctx_ref);
 struct i915_gem_context *
@@ -386,8 +179,7 @@ static inline void i915_gem_context_put(struct i915_gem_context *ctx)
 	kref_put(&ctx->ref, i915_gem_context_release);
 }
 
-void intel_context_init(struct intel_context *ce,
-			struct i915_gem_context *ctx,
-			struct intel_engine_cs *engine);
+struct i915_lut_handle *i915_lut_handle_alloc(void);
+void i915_lut_handle_free(struct i915_lut_handle *lut);
 
 #endif /* !__I915_GEM_CONTEXT_H__ */

drivers/gpu/drm/i915/i915_gem_context_types.h

+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __I915_GEM_CONTEXT_TYPES_H__
+#define __I915_GEM_CONTEXT_TYPES_H__
+
+#include <linux/atomic.h>
+#include <linux/list.h>
+#include <linux/llist.h>
+#include <linux/kref.h>
+#include <linux/mutex.h>
+#include <linux/radix-tree.h>
+#include <linux/rbtree.h>
+#include <linux/rcupdate.h>
+#include <linux/types.h>
+
+#include "i915_scheduler.h"
+#include "intel_context_types.h"
+
+struct pid;
+
+struct drm_i915_private;
+struct drm_i915_file_private;
+struct i915_hw_ppgtt;
+struct i915_timeline;
+struct intel_ring;
+
+/**
+ * struct i915_gem_context - client state
+ *
+ * The struct i915_gem_context represents the combined view of the driver and
+ * logical hardware state for a particular client.
+ */
+struct i915_gem_context {
+	/** i915: i915 device backpointer */
+	struct drm_i915_private *i915;
+
+	/** file_priv: owning file descriptor */
+	struct drm_i915_file_private *file_priv;
+
+	/**
+	 * @ppgtt: unique address space (GTT)
+	 *
+	 * In full-ppgtt mode, each context has its own address space ensuring
+	 * complete seperation of one client from all others.
+	 *
+	 * In other modes, this is a NULL pointer with the expectation that
+	 * the caller uses the shared global GTT.
+	 */
+	struct i915_hw_ppgtt *ppgtt;
+
+	/**
+	 * @pid: process id of creator
+	 *
+	 * Note that who created the context may not be the principle user,
+	 * as the context may be shared across a local socket. However,
+	 * that should only affect the default context, all contexts created
+	 * explicitly by the client are expected to be isolated.
+	 */
+	struct pid *pid;
+
+	/**
+	 * @name: arbitrary name
+	 *
+	 * A name is constructed for the context from the creator's process
+	 * name, pid and user handle in order to uniquely identify the
+	 * context in messages.
+	 */
+	const char *name;
+
+	/** link: place with &drm_i915_private.context_list */
+	struct list_head link;
+	struct llist_node free_link;
+
+	/**
+	 * @ref: reference count
+	 *
+	 * A reference to a context is held by both the client who created it
+	 * and on each request submitted to the hardware using the request
+	 * (to ensure the hardware has access to the state until it has
+	 * finished all pending writes). See i915_gem_context_get() and
+	 * i915_gem_context_put() for access.
+	 */
+	struct kref ref;
+
+	/**
+	 * @rcu: rcu_head for deferred freeing.
+	 */
+	struct rcu_head rcu;
+
+	/**
+	 * @user_flags: small set of booleans controlled by the user
+	 */
+	unsigned long user_flags;
+#define UCONTEXT_NO_ZEROMAP		0
+#define UCONTEXT_NO_ERROR_CAPTURE	1
+#define UCONTEXT_BANNABLE		2
+#define UCONTEXT_RECOVERABLE		3
+
+	/**
+	 * @flags: small set of booleans
+	 */
+	unsigned long flags;
+#define CONTEXT_BANNED			0
+#define CONTEXT_CLOSED			1
+#define CONTEXT_FORCE_SINGLE_SUBMISSION	2
+
+	/**
+	 * @hw_id: - unique identifier for the context
+	 *
+	 * The hardware needs to uniquely identify the context for a few
+	 * functions like fault reporting, PASID, scheduling. The
+	 * &drm_i915_private.context_hw_ida is used to assign a unqiue
+	 * id for the lifetime of the context.
+	 *
+	 * @hw_id_pin_count: - number of times this context had been pinned
+	 * for use (should be, at most, once per engine).
+	 *
+	 * @hw_id_link: - all contexts with an assigned id are tracked
+	 * for possible repossession.
+	 */
+	unsigned int hw_id;
+	atomic_t hw_id_pin_count;
+	struct list_head hw_id_link;
+
+	struct list_head active_engines;
+	struct mutex mutex;
+
+	/**
+	 * @user_handle: userspace identifier
+	 *
+	 * A unique per-file identifier is generated from
+	 * &drm_i915_file_private.contexts.
+	 */
+	u32 user_handle;
+#define DEFAULT_CONTEXT_HANDLE 0
+
+	struct i915_sched_attr sched;
+
+	/** hw_contexts: per-engine logical HW state */
+	struct rb_root hw_contexts;
+	spinlock_t hw_contexts_lock;
+
+	/** ring_size: size for allocating the per-engine ring buffer */
+	u32 ring_size;
+	/** desc_template: invariant fields for the HW context descriptor */
+	u32 desc_template;
+
+	/** guilty_count: How many times this context has caused a GPU hang. */
+	atomic_t guilty_count;
+	/**
+	 * @active_count: How many times this context was active during a GPU
+	 * hang, but did not cause it.
+	 */
+	atomic_t active_count;
+
+	/**
+	 * @hang_timestamp: The last time(s) this context caused a GPU hang
+	 */
+	unsigned long hang_timestamp[2];
+#define CONTEXT_FAST_HANG_JIFFIES (120 * HZ) /* 3 hangs within 120s? Banned! */
+
+	/** remap_slice: Bitmask of cache lines that need remapping */
+	u8 remap_slice;
+
+	/** handles_vma: rbtree to look up our context specific obj/vma for
+	 * the user handle. (user handles are per fd, but the binding is
+	 * per vm, which may be one per context or shared with the global GTT)
+	 */
+	struct radix_tree_root handles_vma;
+
+	/** handles_list: reverse list of all the rbtree entries in use for
+	 * this context, which allows us to free all the allocations on
+	 * context close.
+	 */
+	struct list_head handles_list;
+};
+
+#endif /* __I915_GEM_CONTEXT_TYPES_H__ */

drivers/gpu/drm/i915/i915_gem_dmabuf.c

@@ -300,7 +300,7 @@ struct drm_gem_object *i915_gem_prime_import(struct drm_device *dev,
 
 	get_dma_buf(dma_buf);
 
-	obj = i915_gem_object_alloc(to_i915(dev));
+	obj = i915_gem_object_alloc();
 	if (obj == NULL) {
 		ret = -ENOMEM;
 		goto fail_detach;

drivers/gpu/drm/i915/i915_gem_evict.c

@@ -38,31 +38,21 @@ I915_SELFTEST_DECLARE(static struct igt_evict_ctl {
 
 static bool ggtt_is_idle(struct drm_i915_private *i915)
 {
-       struct intel_engine_cs *engine;
-       enum intel_engine_id id;
-
-       if (i915->gt.active_requests)
-	       return false;
-
-       for_each_engine(engine, i915, id) {
-	       if (!intel_engine_has_kernel_context(engine))
-		       return false;
-       }
-
-       return true;
+	return !i915->gt.active_requests;
 }
 
 static int ggtt_flush(struct drm_i915_private *i915)
 {
 	int err;
 
-	/* Not everything in the GGTT is tracked via vma (otherwise we
+	/*
+	 * Not everything in the GGTT is tracked via vma (otherwise we
 	 * could evict as required with minimal stalling) so we are forced
 	 * to idle the GPU and explicitly retire outstanding requests in
 	 * the hopes that we can then remove contexts and the like only
 	 * bound by their active reference.
 	 */
-	err = i915_gem_switch_to_kernel_context(i915);
+	err = i915_gem_switch_to_kernel_context(i915, i915->gt.active_engines);
 	if (err)
 		return err;
 

drivers/gpu/drm/i915/i915_gem_execbuffer.c

@@ -794,8 +794,8 @@ static int eb_wait_for_ring(const struct i915_execbuffer *eb)
 	 * keeping all of their resources pinned.
 	 */
 
-	ce = to_intel_context(eb->ctx, eb->engine);
-	if (!ce->ring) /* first use, assume empty! */
+	ce = intel_context_lookup(eb->ctx, eb->engine);
+	if (!ce || !ce->ring) /* first use, assume empty! */
 		return 0;
 
 	rq = __eb_wait_for_ring(ce->ring);
@@ -849,12 +849,12 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 		}
 
 		vma = i915_vma_instance(obj, eb->vm, NULL);
-		if (unlikely(IS_ERR(vma))) {
+		if (IS_ERR(vma)) {
 			err = PTR_ERR(vma);
 			goto err_obj;
 		}
 
-		lut = kmem_cache_alloc(eb->i915->luts, GFP_KERNEL);
+		lut = i915_lut_handle_alloc();
 		if (unlikely(!lut)) {
 			err = -ENOMEM;
 			goto err_obj;
@@ -862,7 +862,7 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 
 		err = radix_tree_insert(handles_vma, handle, vma);
 		if (unlikely(err)) {
-			kmem_cache_free(eb->i915->luts, lut);
+			i915_lut_handle_free(lut);
 			goto err_obj;
 		}
 
@@ -1957,7 +1957,7 @@ static int i915_reset_gen7_sol_offsets(struct i915_request *rq)
 	u32 *cs;
 	int i;
 
-	if (!IS_GEN(rq->i915, 7) || rq->engine->id != RCS) {
+	if (!IS_GEN(rq->i915, 7) || rq->engine->id != RCS0) {
 		DRM_DEBUG("sol reset is gen7/rcs only\n");
 		return -EINVAL;
 	}
@@ -2082,11 +2082,11 @@ gen8_dispatch_bsd_engine(struct drm_i915_private *dev_priv,
 #define I915_USER_RINGS (4)
 
 static const enum intel_engine_id user_ring_map[I915_USER_RINGS + 1] = {
-	[I915_EXEC_DEFAULT]	= RCS,
-	[I915_EXEC_RENDER]	= RCS,
-	[I915_EXEC_BLT]		= BCS,
-	[I915_EXEC_BSD]		= VCS,
-	[I915_EXEC_VEBOX]	= VECS
+	[I915_EXEC_DEFAULT]	= RCS0,
+	[I915_EXEC_RENDER]	= RCS0,
+	[I915_EXEC_BLT]		= BCS0,
+	[I915_EXEC_BSD]		= VCS0,
+	[I915_EXEC_VEBOX]	= VECS0
 };
 
 static struct intel_engine_cs *
@@ -2109,7 +2109,7 @@ eb_select_engine(struct drm_i915_private *dev_priv,
 		return NULL;
 	}
 
-	if (user_ring_id == I915_EXEC_BSD && HAS_BSD2(dev_priv)) {
+	if (user_ring_id == I915_EXEC_BSD && HAS_ENGINE(dev_priv, VCS1)) {
 		unsigned int bsd_idx = args->flags & I915_EXEC_BSD_MASK;
 
 		if (bsd_idx == I915_EXEC_BSD_DEFAULT) {
@@ -2312,10 +2312,6 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	if (args->flags & I915_EXEC_IS_PINNED)
 		eb.batch_flags |= I915_DISPATCH_PINNED;
 
-	eb.engine = eb_select_engine(eb.i915, file, args);
-	if (!eb.engine)
-		return -EINVAL;
-
 	if (args->flags & I915_EXEC_FENCE_IN) {
 		in_fence = sync_file_get_fence(lower_32_bits(args->rsvd2));
 		if (!in_fence)
@@ -2340,6 +2336,12 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	if (unlikely(err))
 		goto err_destroy;
 
+	eb.engine = eb_select_engine(eb.i915, file, args);
+	if (!eb.engine) {
+		err = -EINVAL;
+		goto err_engine;
+	}
+
 	/*
 	 * Take a local wakeref for preparing to dispatch the execbuf as
 	 * we expect to access the hardware fairly frequently in the
@@ -2505,6 +2507,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	mutex_unlock(&dev->struct_mutex);
 err_rpm:
 	intel_runtime_pm_put(eb.i915, wakeref);
+err_engine:
 	i915_gem_context_put(eb.ctx);
 err_destroy:
 	eb_destroy(&eb);

drivers/gpu/drm/i915/i915_gem_fence_reg.c

@@ -210,6 +210,7 @@ static int fence_update(struct drm_i915_fence_reg *fence,
 			struct i915_vma *vma)
 {
 	intel_wakeref_t wakeref;
+	struct i915_vma *old;
 	int ret;
 
 	if (vma) {
@@ -229,49 +230,55 @@ static int fence_update(struct drm_i915_fence_reg *fence,
 			return ret;
 	}
 
-	if (fence->vma) {
-		struct i915_vma *old = fence->vma;
-
+	old = xchg(&fence->vma, NULL);
+	if (old) {
 		ret = i915_active_request_retire(&old->last_fence,
 					     &old->obj->base.dev->struct_mutex);
-		if (ret)
+		if (ret) {
+			fence->vma = old;
 			return ret;
+		}
 
 		i915_vma_flush_writes(old);
-	}
 
-	if (fence->vma && fence->vma != vma) {
-		/* Ensure that all userspace CPU access is completed before
+		/*
+		 * Ensure that all userspace CPU access is completed before
 		 * stealing the fence.
 		 */
-		GEM_BUG_ON(fence->vma->fence != fence);
-		i915_vma_revoke_mmap(fence->vma);
-
-		fence->vma->fence = NULL;
-		fence->vma = NULL;
+		if (old != vma) {
+			GEM_BUG_ON(old->fence != fence);
+			i915_vma_revoke_mmap(old);
+			old->fence = NULL;
+		}
 
 		list_move(&fence->link, &fence->i915->mm.fence_list);
 	}
 
-	/* We only need to update the register itself if the device is awake.
+	/*
+	 * We only need to update the register itself if the device is awake.
 	 * If the device is currently powered down, we will defer the write
 	 * to the runtime resume, see i915_gem_restore_fences().
+	 *
+	 * This only works for removing the fence register, on acquisition
+	 * the caller must hold the rpm wakeref. The fence register must
+	 * be cleared before we can use any other fences to ensure that
+	 * the new fences do not overlap the elided clears, confusing HW.
 	 */
 	wakeref = intel_runtime_pm_get_if_in_use(fence->i915);
-	if (wakeref) {
-		fence_write(fence, vma);
-		intel_runtime_pm_put(fence->i915, wakeref);
+	if (!wakeref) {
+		GEM_BUG_ON(vma);
+		return 0;
 	}
 
-	if (vma) {
-		if (fence->vma != vma) {
-			vma->fence = fence;
-			fence->vma = vma;
-		}
+	WRITE_ONCE(fence->vma, vma);
+	fence_write(fence, vma);
 
+	if (vma) {
+		vma->fence = fence;
 		list_move_tail(&fence->link, &fence->i915->mm.fence_list);
 	}
 
+	intel_runtime_pm_put(fence->i915, wakeref);
 	return 0;
 }
 
@@ -435,32 +442,6 @@ void i915_unreserve_fence(struct drm_i915_fence_reg *fence)
 	list_add(&fence->link, &fence->i915->mm.fence_list);
 }
 
-/**
- * i915_gem_revoke_fences - revoke fence state
- * @dev_priv: i915 device private
- *
- * Removes all GTT mmappings via the fence registers. This forces any user
- * of the fence to reacquire that fence before continuing with their access.
- * One use is during GPU reset where the fence register is lost and we need to
- * revoke concurrent userspace access via GTT mmaps until the hardware has been
- * reset and the fence registers have been restored.
- */
-void i915_gem_revoke_fences(struct drm_i915_private *dev_priv)
-{
-	int i;
-
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
-	for (i = 0; i < dev_priv->num_fence_regs; i++) {
-		struct drm_i915_fence_reg *fence = &dev_priv->fence_regs[i];
-
-		GEM_BUG_ON(fence->vma && fence->vma->fence != fence);
-
-		if (fence->vma)
-			i915_vma_revoke_mmap(fence->vma);
-	}
-}
-
 /**
  * i915_gem_restore_fences - restore fence state
  * @dev_priv: i915 device private
@@ -473,9 +454,10 @@ void i915_gem_restore_fences(struct drm_i915_private *dev_priv)
 {
 	int i;
 
+	rcu_read_lock(); /* keep obj alive as we dereference */
 	for (i = 0; i < dev_priv->num_fence_regs; i++) {
 		struct drm_i915_fence_reg *reg = &dev_priv->fence_regs[i];
-		struct i915_vma *vma = reg->vma;
+		struct i915_vma *vma = READ_ONCE(reg->vma);
 
 		GEM_BUG_ON(vma && vma->fence != reg);
 
@@ -483,18 +465,12 @@ void i915_gem_restore_fences(struct drm_i915_private *dev_priv)
 		 * Commit delayed tiling changes if we have an object still
 		 * attached to the fence, otherwise just clear the fence.
 		 */
-		if (vma && !i915_gem_object_is_tiled(vma->obj)) {
-			GEM_BUG_ON(!reg->dirty);
-			GEM_BUG_ON(i915_vma_has_userfault(vma));
-
-			list_move(&reg->link, &dev_priv->mm.fence_list);
-			vma->fence = NULL;
+		if (vma && !i915_gem_object_is_tiled(vma->obj))
 			vma = NULL;
-		}
 
 		fence_write(reg, vma);
-		reg->vma = vma;
 	}
+	rcu_read_unlock();
 }
 
 /**

drivers/gpu/drm/i915/i915_gem_gtt.c

@@ -584,7 +584,7 @@ setup_scratch_page(struct i915_address_space *vm, gfp_t gfp)
 	 * for all.
 	 */
 	size = I915_GTT_PAGE_SIZE_4K;
-	if (i915_vm_is_48bit(vm) &&
+	if (i915_vm_is_4lvl(vm) &&
 	    HAS_PAGE_SIZES(vm->i915, I915_GTT_PAGE_SIZE_64K)) {
 		size = I915_GTT_PAGE_SIZE_64K;
 		gfp |= __GFP_NOWARN;
@@ -613,7 +613,7 @@ setup_scratch_page(struct i915_address_space *vm, gfp_t gfp)
 
 		vm->scratch_page.page = page;
 		vm->scratch_page.daddr = addr;
-		vm->scratch_page.order = order;
+		vm->scratch_order = order;
 		return 0;
 
 unmap_page:
@@ -632,10 +632,11 @@ setup_scratch_page(struct i915_address_space *vm, gfp_t gfp)
 static void cleanup_scratch_page(struct i915_address_space *vm)
 {
 	struct i915_page_dma *p = &vm->scratch_page;
+	int order = vm->scratch_order;
 
-	dma_unmap_page(vm->dma, p->daddr, BIT(p->order) << PAGE_SHIFT,
+	dma_unmap_page(vm->dma, p->daddr, BIT(order) << PAGE_SHIFT,
 		       PCI_DMA_BIDIRECTIONAL);
-	__free_pages(p->page, p->order);
+	__free_pages(p->page, order);
 }
 
 static struct i915_page_table *alloc_pt(struct i915_address_space *vm)
@@ -726,18 +727,13 @@ static void __pdp_fini(struct i915_page_directory_pointer *pdp)
 	pdp->page_directory = NULL;
 }
 
-static inline bool use_4lvl(const struct i915_address_space *vm)
-{
-	return i915_vm_is_48bit(vm);
-}
-
 static struct i915_page_directory_pointer *
 alloc_pdp(struct i915_address_space *vm)
 {
 	struct i915_page_directory_pointer *pdp;
 	int ret = -ENOMEM;
 
-	GEM_BUG_ON(!use_4lvl(vm));
+	GEM_BUG_ON(!i915_vm_is_4lvl(vm));
 
 	pdp = kzalloc(sizeof(*pdp), GFP_KERNEL);
 	if (!pdp)
@@ -766,7 +762,7 @@ static void free_pdp(struct i915_address_space *vm,
 {
 	__pdp_fini(pdp);
 
-	if (!use_4lvl(vm))
+	if (!i915_vm_is_4lvl(vm))
 		return;
 
 	cleanup_px(vm, pdp);
@@ -791,14 +787,15 @@ static void gen8_initialize_pml4(struct i915_address_space *vm,
 	memset_p((void **)pml4->pdps, vm->scratch_pdp, GEN8_PML4ES_PER_PML4);
 }
 
-/* PDE TLBs are a pain to invalidate on GEN8+. When we modify
+/*
+ * PDE TLBs are a pain to invalidate on GEN8+. When we modify
  * the page table structures, we mark them dirty so that
  * context switching/execlist queuing code takes extra steps
  * to ensure that tlbs are flushed.
  */
 static void mark_tlbs_dirty(struct i915_hw_ppgtt *ppgtt)
 {
-	ppgtt->pd_dirty_rings = INTEL_INFO(ppgtt->vm.i915)->ring_mask;
+	ppgtt->pd_dirty_engines = ALL_ENGINES;
 }
 
 /* Removes entries from a single page table, releasing it if it's empty.
@@ -809,8 +806,6 @@ static bool gen8_ppgtt_clear_pt(const struct i915_address_space *vm,
 				u64 start, u64 length)
 {
 	unsigned int num_entries = gen8_pte_count(start, length);
-	unsigned int pte = gen8_pte_index(start);
-	unsigned int pte_end = pte + num_entries;
 	gen8_pte_t *vaddr;
 
 	GEM_BUG_ON(num_entries > pt->used_ptes);
@@ -820,8 +815,7 @@ static bool gen8_ppgtt_clear_pt(const struct i915_address_space *vm,
 		return true;
 
 	vaddr = kmap_atomic_px(pt);
-	while (pte < pte_end)
-		vaddr[pte++] = vm->scratch_pte;
+	memset64(vaddr + gen8_pte_index(start), vm->scratch_pte, num_entries);
 	kunmap_atomic(vaddr);
 
 	return false;
@@ -872,7 +866,7 @@ static void gen8_ppgtt_set_pdpe(struct i915_address_space *vm,
 	gen8_ppgtt_pdpe_t *vaddr;
 
 	pdp->page_directory[pdpe] = pd;
-	if (!use_4lvl(vm))
+	if (!i915_vm_is_4lvl(vm))
 		return;
 
 	vaddr = kmap_atomic_px(pdp);
@@ -937,7 +931,7 @@ static void gen8_ppgtt_clear_4lvl(struct i915_address_space *vm,
 	struct i915_page_directory_pointer *pdp;
 	unsigned int pml4e;
 
-	GEM_BUG_ON(!use_4lvl(vm));
+	GEM_BUG_ON(!i915_vm_is_4lvl(vm));
 
 	gen8_for_each_pml4e(pdp, pml4, start, length, pml4e) {
 		GEM_BUG_ON(pdp == vm->scratch_pdp);
@@ -1219,7 +1213,7 @@ static int gen8_init_scratch(struct i915_address_space *vm)
 
 		GEM_BUG_ON(!clone->has_read_only);
 
-		vm->scratch_page.order = clone->scratch_page.order;
+		vm->scratch_order = clone->scratch_order;
 		vm->scratch_pte = clone->scratch_pte;
 		vm->scratch_pt  = clone->scratch_pt;
 		vm->scratch_pd  = clone->scratch_pd;
@@ -1248,7 +1242,7 @@ static int gen8_init_scratch(struct i915_address_space *vm)
 		goto free_pt;
 	}
 
-	if (use_4lvl(vm)) {
+	if (i915_vm_is_4lvl(vm)) {
 		vm->scratch_pdp = alloc_pdp(vm);
 		if (IS_ERR(vm->scratch_pdp)) {
 			ret = PTR_ERR(vm->scratch_pdp);
@@ -1258,7 +1252,7 @@ static int gen8_init_scratch(struct i915_address_space *vm)
 
 	gen8_initialize_pt(vm, vm->scratch_pt);
 	gen8_initialize_pd(vm, vm->scratch_pd);
-	if (use_4lvl(vm))
+	if (i915_vm_is_4lvl(vm))
 		gen8_initialize_pdp(vm, vm->scratch_pdp);
 
 	return 0;
@@ -1280,7 +1274,7 @@ static int gen8_ppgtt_notify_vgt(struct i915_hw_ppgtt *ppgtt, bool create)
 	enum vgt_g2v_type msg;
 	int i;
 
-	if (use_4lvl(vm)) {
+	if (i915_vm_is_4lvl(vm)) {
 		const u64 daddr = px_dma(&ppgtt->pml4);
 
 		I915_WRITE(vgtif_reg(pdp[0].lo), lower_32_bits(daddr));
@@ -1310,7 +1304,7 @@ static void gen8_free_scratch(struct i915_address_space *vm)
 	if (!vm->scratch_page.daddr)
 		return;
 
-	if (use_4lvl(vm))
+	if (i915_vm_is_4lvl(vm))
 		free_pdp(vm, vm->scratch_pdp);
 	free_pd(vm, vm->scratch_pd);
 	free_pt(vm, vm->scratch_pt);
@@ -1356,7 +1350,7 @@ static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
 	if (intel_vgpu_active(dev_priv))
 		gen8_ppgtt_notify_vgt(ppgtt, false);
 
-	if (use_4lvl(vm))
+	if (i915_vm_is_4lvl(vm))
 		gen8_ppgtt_cleanup_4lvl(ppgtt);
 	else
 		gen8_ppgtt_cleanup_3lvl(&ppgtt->vm, &ppgtt->pdp);
@@ -1519,6 +1513,23 @@ static int gen8_preallocate_top_level_pdp(struct i915_hw_ppgtt *ppgtt)
 	return -ENOMEM;
 }
 
+static void ppgtt_init(struct drm_i915_private *i915,
+		       struct i915_hw_ppgtt *ppgtt)
+{
+	kref_init(&ppgtt->ref);
+
+	ppgtt->vm.i915 = i915;
+	ppgtt->vm.dma = &i915->drm.pdev->dev;
+	ppgtt->vm.total = BIT_ULL(INTEL_INFO(i915)->ppgtt_size);
+
+	i915_address_space_init(&ppgtt->vm, VM_CLASS_PPGTT);
+
+	ppgtt->vm.vma_ops.bind_vma    = ppgtt_bind_vma;
+	ppgtt->vm.vma_ops.unbind_vma  = ppgtt_unbind_vma;
+	ppgtt->vm.vma_ops.set_pages   = ppgtt_set_pages;
+	ppgtt->vm.vma_ops.clear_pages = clear_pages;
+}
+
 /*
  * GEN8 legacy ppgtt programming is accomplished through a max 4 PDP registers
  * with a net effect resembling a 2-level page table in normal x86 terms. Each
@@ -1535,20 +1546,11 @@ static struct i915_hw_ppgtt *gen8_ppgtt_create(struct drm_i915_private *i915)
 	if (!ppgtt)
 		return ERR_PTR(-ENOMEM);
 
-	kref_init(&ppgtt->ref);
-
-	ppgtt->vm.i915 = i915;
-	ppgtt->vm.dma = &i915->drm.pdev->dev;
-
-	ppgtt->vm.total = HAS_FULL_48BIT_PPGTT(i915) ?
-		1ULL << 48 :
-		1ULL << 32;
+	ppgtt_init(i915, ppgtt);
 
 	/* From bdw, there is support for read-only pages in the PPGTT. */
 	ppgtt->vm.has_read_only = true;
 
-	i915_address_space_init(&ppgtt->vm, VM_CLASS_PPGTT);
-
 	/* There are only few exceptions for gen >=6. chv and bxt.
 	 * And we are not sure about the latter so play safe for now.
 	 */
@@ -1559,7 +1561,7 @@ static struct i915_hw_ppgtt *gen8_ppgtt_create(struct drm_i915_private *i915)
 	if (err)
 		goto err_free;
 
-	if (use_4lvl(&ppgtt->vm)) {
+	if (i915_vm_is_4lvl(&ppgtt->vm)) {
 		err = setup_px(&ppgtt->vm, &ppgtt->pml4);
 		if (err)
 			goto err_scratch;
@@ -1592,11 +1594,6 @@ static struct i915_hw_ppgtt *gen8_ppgtt_create(struct drm_i915_private *i915)
 
 	ppgtt->vm.cleanup = gen8_ppgtt_cleanup;
 
-	ppgtt->vm.vma_ops.bind_vma    = ppgtt_bind_vma;
-	ppgtt->vm.vma_ops.unbind_vma  = ppgtt_unbind_vma;
-	ppgtt->vm.vma_ops.set_pages   = ppgtt_set_pages;
-	ppgtt->vm.vma_ops.clear_pages = clear_pages;
-
 	return ppgtt;
 
 err_scratch:
@@ -1672,8 +1669,7 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm,
 
 	while (num_entries) {
 		struct i915_page_table *pt = ppgtt->base.pd.page_table[pde++];
-		const unsigned int end = min(pte + num_entries, GEN6_PTES);
-		const unsigned int count = end - pte;
+		const unsigned int count = min(num_entries, GEN6_PTES - pte);
 		gen6_pte_t *vaddr;
 
 		GEM_BUG_ON(pt == vm->scratch_pt);
@@ -1693,9 +1689,7 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm,
 		 */
 
 		vaddr = kmap_atomic_px(pt);
-		do {
-			vaddr[pte++] = scratch_pte;
-		} while (pte < end);
+		memset32(vaddr + pte, scratch_pte, count);
 		kunmap_atomic(vaddr);
 
 		pte = 0;
@@ -1913,7 +1907,7 @@ static struct i915_vma *pd_vma_create(struct gen6_hw_ppgtt *ppgtt, int size)
 	GEM_BUG_ON(!IS_ALIGNED(size, I915_GTT_PAGE_SIZE));
 	GEM_BUG_ON(size > ggtt->vm.total);
 
-	vma = kmem_cache_zalloc(i915->vmas, GFP_KERNEL);
+	vma = i915_vma_alloc();
 	if (!vma)
 		return ERR_PTR(-ENOMEM);
 
@@ -1991,25 +1985,13 @@ static struct i915_hw_ppgtt *gen6_ppgtt_create(struct drm_i915_private *i915)
 	if (!ppgtt)
 		return ERR_PTR(-ENOMEM);
 
-	kref_init(&ppgtt->base.ref);
-
-	ppgtt->base.vm.i915 = i915;
-	ppgtt->base.vm.dma = &i915->drm.pdev->dev;
-
-	ppgtt->base.vm.total = I915_PDES * GEN6_PTES * I915_GTT_PAGE_SIZE;
-
-	i915_address_space_init(&ppgtt->base.vm, VM_CLASS_PPGTT);
+	ppgtt_init(i915, &ppgtt->base);
 
 	ppgtt->base.vm.allocate_va_range = gen6_alloc_va_range;
 	ppgtt->base.vm.clear_range = gen6_ppgtt_clear_range;
 	ppgtt->base.vm.insert_entries = gen6_ppgtt_insert_entries;
 	ppgtt->base.vm.cleanup = gen6_ppgtt_cleanup;
 
-	ppgtt->base.vm.vma_ops.bind_vma    = ppgtt_bind_vma;
-	ppgtt->base.vm.vma_ops.unbind_vma  = ppgtt_unbind_vma;
-	ppgtt->base.vm.vma_ops.set_pages   = ppgtt_set_pages;
-	ppgtt->base.vm.vma_ops.clear_pages = clear_pages;
-
 	ppgtt->base.vm.pte_encode = ggtt->vm.pte_encode;
 
 	err = gen6_ppgtt_init_scratch(ppgtt);
@@ -3701,7 +3683,7 @@ i915_get_ggtt_vma_pages(struct i915_vma *vma)
 	}
 
 	ret = 0;
-	if (unlikely(IS_ERR(vma->pages))) {
+	if (IS_ERR(vma->pages)) {
 		ret = PTR_ERR(vma->pages);
 		vma->pages = NULL;
 		DRM_ERROR("Failed to get pages for VMA view type %u (%d)!\n",

drivers/gpu/drm/i915/i915_gem_gtt.h

@@ -213,7 +213,6 @@ struct i915_vma;
 
 struct i915_page_dma {
 	struct page *page;
-	int order;
 	union {
 		dma_addr_t daddr;
 
@@ -293,6 +292,7 @@ struct i915_address_space {
 #define VM_CLASS_PPGTT 1
 
 	u64 scratch_pte;
+	int scratch_order;
 	struct i915_page_dma scratch_page;
 	struct i915_page_table *scratch_pt;
 	struct i915_page_directory *scratch_pd;
@@ -348,7 +348,7 @@ struct i915_address_space {
 #define i915_is_ggtt(vm) ((vm)->is_ggtt)
 
 static inline bool
-i915_vm_is_48bit(const struct i915_address_space *vm)
+i915_vm_is_4lvl(const struct i915_address_space *vm)
 {
 	return (vm->total - 1) >> 32;
 }
@@ -356,7 +356,7 @@ i915_vm_is_48bit(const struct i915_address_space *vm)
 static inline bool
 i915_vm_has_scratch_64K(struct i915_address_space *vm)
 {
-	return vm->scratch_page.order == get_order(I915_GTT_PAGE_SIZE_64K);
+	return vm->scratch_order == get_order(I915_GTT_PAGE_SIZE_64K);
 }
 
 /* The Graphics Translation Table is the way in which GEN hardware translates a
@@ -390,7 +390,7 @@ struct i915_hw_ppgtt {
 	struct i915_address_space vm;
 	struct kref ref;
 
-	unsigned long pd_dirty_rings;
+	unsigned long pd_dirty_engines;
 	union {
 		struct i915_pml4 pml4;		/* GEN8+ & 48b PPGTT */
 		struct i915_page_directory_pointer pdp;	/* GEN8+ */
@@ -488,7 +488,7 @@ static inline u32 gen6_pde_index(u32 addr)
 static inline unsigned int
 i915_pdpes_per_pdp(const struct i915_address_space *vm)
 {
-	if (i915_vm_is_48bit(vm))
+	if (i915_vm_is_4lvl(vm))
 		return GEN8_PML4ES_PER_PML4;
 
 	return GEN8_3LVL_PDPES;

drivers/gpu/drm/i915/i915_gem_internal.c

@@ -193,7 +193,7 @@ i915_gem_object_create_internal(struct drm_i915_private *i915,
 	if (overflows_type(size, obj->base.size))
 		return ERR_PTR(-E2BIG);
 
-	obj = i915_gem_object_alloc(i915);
+	obj = i915_gem_object_alloc();
 	if (!obj)
 		return ERR_PTR(-ENOMEM);
 

drivers/gpu/drm/i915/i915_gem_object.c

@@ -24,6 +24,22 @@
 
 #include "i915_drv.h"
 #include "i915_gem_object.h"
+#include "i915_globals.h"
+
+static struct i915_global_object {
+	struct i915_global base;
+	struct kmem_cache *slab_objects;
+} global;
+
+struct drm_i915_gem_object *i915_gem_object_alloc(void)
+{
+	return kmem_cache_zalloc(global.slab_objects, GFP_KERNEL);
+}
+
+void i915_gem_object_free(struct drm_i915_gem_object *obj)
+{
+	return kmem_cache_free(global.slab_objects, obj);
+}
 
 /**
  * Mark up the object's coherency levels for a given cache_level
@@ -46,3 +62,29 @@ void i915_gem_object_set_cache_coherency(struct drm_i915_gem_object *obj,
 	obj->cache_dirty =
 		!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE);
 }
+
+static void i915_global_objects_shrink(void)
+{
+	kmem_cache_shrink(global.slab_objects);
+}
+
+static void i915_global_objects_exit(void)
+{
+	kmem_cache_destroy(global.slab_objects);
+}
+
+static struct i915_global_object global = { {
+	.shrink = i915_global_objects_shrink,
+	.exit = i915_global_objects_exit,
+} };
+
+int __init i915_global_objects_init(void)
+{
+	global.slab_objects =
+		KMEM_CACHE(drm_i915_gem_object, SLAB_HWCACHE_ALIGN);
+	if (!global.slab_objects)
+		return -ENOMEM;
+
+	i915_global_register(&global.base);
+	return 0;
+}

drivers/gpu/drm/i915/i915_gem_object.h

@@ -304,6 +304,9 @@ to_intel_bo(struct drm_gem_object *gem)
 	return container_of(gem, struct drm_i915_gem_object, base);
 }
 
+struct drm_i915_gem_object *i915_gem_object_alloc(void);
+void i915_gem_object_free(struct drm_i915_gem_object *obj);
+
 /**
  * i915_gem_object_lookup_rcu - look up a temporary GEM object from its handle
  * @filp: DRM file private date
@@ -500,4 +503,3 @@ void i915_gem_object_set_cache_coherency(struct drm_i915_gem_object *obj,
 void i915_gem_object_flush_if_display(struct drm_i915_gem_object *obj);
 
 #endif
-

drivers/gpu/drm/i915/i915_gem_render_state.c

@@ -42,7 +42,7 @@ struct intel_render_state {
 static const struct intel_renderstate_rodata *
 render_state_get_rodata(const struct intel_engine_cs *engine)
 {
-	if (engine->id != RCS)
+	if (engine->id != RCS0)
 		return NULL;
 
 	switch (INTEL_GEN(engine->i915)) {

drivers/gpu/drm/i915/i915_gem_stolen.c

@@ -565,7 +565,7 @@ _i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
 	struct drm_i915_gem_object *obj;
 	unsigned int cache_level;
 
-	obj = i915_gem_object_alloc(dev_priv);
+	obj = i915_gem_object_alloc();
 	if (obj == NULL)
 		return NULL;
 

drivers/gpu/drm/i915/i915_gem_userptr.c

@@ -795,7 +795,7 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 			return -ENODEV;
 	}
 
-	obj = i915_gem_object_alloc(dev_priv);
+	obj = i915_gem_object_alloc();
 	if (obj == NULL)
 		return -ENOMEM;
 

drivers/gpu/drm/i915/i915_globals.c

+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/slab.h>
+#include <linux/workqueue.h>
+
+#include "i915_active.h"
+#include "i915_gem_context.h"
+#include "i915_gem_object.h"
+#include "i915_globals.h"
+#include "i915_request.h"
+#include "i915_scheduler.h"
+#include "i915_vma.h"
+
+static LIST_HEAD(globals);
+
+void __init i915_global_register(struct i915_global *global)
+{
+	GEM_BUG_ON(!global->shrink);
+	GEM_BUG_ON(!global->exit);
+
+	list_add_tail(&global->link, &globals);
+}
+
+static void __i915_globals_cleanup(void)
+{
+	struct i915_global *global, *next;
+
+	list_for_each_entry_safe_reverse(global, next, &globals, link)
+		global->exit();
+}
+
+static __initconst int (* const initfn[])(void) = {
+	i915_global_active_init,
+	i915_global_context_init,
+	i915_global_gem_context_init,
+	i915_global_objects_init,
+	i915_global_request_init,
+	i915_global_scheduler_init,
+	i915_global_vma_init,
+};
+
+int __init i915_globals_init(void)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(initfn); i++) {
+		int err;
+
+		err = initfn[i]();
+		if (err) {
+			__i915_globals_cleanup();
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+static void i915_globals_shrink(void)
+{
+	struct i915_global *global;
+
+	/*
+	 * kmem_cache_shrink() discards empty slabs and reorders partially
+	 * filled slabs to prioritise allocating from the mostly full slabs,
+	 * with the aim of reducing fragmentation.
+	 */
+	list_for_each_entry(global, &globals, link)
+		global->shrink();
+}
+
+static atomic_t active;
+static atomic_t epoch;
+struct park_work {
+	struct rcu_work work;
+	int epoch;
+};
+
+static void __i915_globals_park(struct work_struct *work)
+{
+	struct park_work *wrk = container_of(work, typeof(*wrk), work.work);
+
+	/* Confirm nothing woke up in the last grace period */
+	if (wrk->epoch == atomic_read(&epoch))
+		i915_globals_shrink();
+
+	kfree(wrk);
+}
+
+void i915_globals_park(void)
+{
+	struct park_work *wrk;
+
+	/*
+	 * Defer shrinking the global slab caches (and other work) until
+	 * after a RCU grace period has completed with no activity. This
+	 * is to try and reduce the latency impact on the consumers caused
+	 * by us shrinking the caches the same time as they are trying to
+	 * allocate, with the assumption being that if we idle long enough
+	 * for an RCU grace period to elapse since the last use, it is likely
+	 * to be longer until we need the caches again.
+	 */
+	if (!atomic_dec_and_test(&active))
+		return;
+
+	wrk = kmalloc(sizeof(*wrk), GFP_KERNEL);
+	if (!wrk)
+		return;
+
+	wrk->epoch = atomic_inc_return(&epoch);
+	INIT_RCU_WORK(&wrk->work, __i915_globals_park);
+	queue_rcu_work(system_wq, &wrk->work);
+}
+
+void i915_globals_unpark(void)
+{
+	atomic_inc(&epoch);
+	atomic_inc(&active);
+}
+
+void __exit i915_globals_exit(void)
+{
+	/* Flush any residual park_work */
+	rcu_barrier();
+	flush_scheduled_work();
+
+	__i915_globals_cleanup();
+
+	/* And ensure that our DESTROY_BY_RCU slabs are truly destroyed */
+	rcu_barrier();
+}

drivers/gpu/drm/i915/i915_globals.h

+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef _I915_GLOBALS_H_
+#define _I915_GLOBALS_H_
+
+typedef void (*i915_global_func_t)(void);
+
+struct i915_global {
+	struct list_head link;
+
+	i915_global_func_t shrink;
+	i915_global_func_t exit;
+};
+
+void i915_global_register(struct i915_global *global);
+
+int i915_globals_init(void);
+void i915_globals_park(void);
+void i915_globals_unpark(void);
+void i915_globals_exit(void);
+
+/* constructors */
+int i915_global_active_init(void);
+int i915_global_context_init(void);
+int i915_global_gem_context_init(void);
+int i915_global_objects_init(void);
+int i915_global_request_init(void);
+int i915_global_scheduler_init(void);
+int i915_global_vma_init(void);
+
+#endif /* _I915_GLOBALS_H_ */

drivers/gpu/drm/i915/i915_pvinfo.h

@@ -52,7 +52,7 @@ enum vgt_g2v_type {
 /*
  * VGT capabilities type
  */
-#define VGT_CAPS_FULL_48BIT_PPGTT	BIT(2)
+#define VGT_CAPS_FULL_PPGTT		BIT(2)
 #define VGT_CAPS_HWSP_EMULATION		BIT(3)
 #define VGT_CAPS_HUGE_GTT		BIT(4)
 

drivers/gpu/drm/i915/i915_query.c

@@ -10,12 +10,34 @@
 #include "i915_query.h"
 #include <uapi/drm/i915_drm.h>
 
+static int copy_query_item(void *query_hdr, size_t query_sz,
+			   u32 total_length,
+			   struct drm_i915_query_item *query_item)
+{
+	if (query_item->length == 0)
+		return total_length;
+
+	if (query_item->length < total_length)
+		return -EINVAL;
+
+	if (copy_from_user(query_hdr, u64_to_user_ptr(query_item->data_ptr),
+			   query_sz))
+		return -EFAULT;
+
+	if (!access_ok(u64_to_user_ptr(query_item->data_ptr),
+		       total_length))
+		return -EFAULT;
+
+	return 0;
+}
+
 static int query_topology_info(struct drm_i915_private *dev_priv,
 			       struct drm_i915_query_item *query_item)
 {
 	const struct sseu_dev_info *sseu = &RUNTIME_INFO(dev_priv)->sseu;
 	struct drm_i915_query_topology_info topo;
 	u32 slice_length, subslice_length, eu_length, total_length;
+	int ret;
 
 	if (query_item->flags != 0)
 		return -EINVAL;
@@ -33,23 +55,14 @@ static int query_topology_info(struct drm_i915_private *dev_priv,
 
 	total_length = sizeof(topo) + slice_length + subslice_length + eu_length;
 
-	if (query_item->length == 0)
-		return total_length;
-
-	if (query_item->length < total_length)
-		return -EINVAL;
-
-	if (copy_from_user(&topo, u64_to_user_ptr(query_item->data_ptr),
-			   sizeof(topo)))
-		return -EFAULT;
+	ret = copy_query_item(&topo, sizeof(topo), total_length,
+			      query_item);
+	if (ret != 0)
+		return ret;
 
 	if (topo.flags != 0)
 		return -EINVAL;
 
-	if (!access_ok(u64_to_user_ptr(query_item->data_ptr),
-		       total_length))
-		return -EFAULT;
-
 	memset(&topo, 0, sizeof(topo));
 	topo.max_slices = sseu->max_slices;
 	topo.max_subslices = sseu->max_subslices;

drivers/gpu/drm/i915/i915_sw_fence.h

@@ -79,6 +79,9 @@ int i915_sw_fence_await_reservation(struct i915_sw_fence *fence,
 				    unsigned long timeout,
 				    gfp_t gfp);
 
+void i915_sw_fence_await(struct i915_sw_fence *fence);
+void i915_sw_fence_complete(struct i915_sw_fence *fence);
+
 static inline bool i915_sw_fence_signaled(const struct i915_sw_fence *fence)
 {
 	return atomic_read(&fence->pending) <= 0;