Merge tag 'drm-intel-next-2017-06-19' of git://anongit.freedesktop.org/git/drm-intel into drm-next

Final pile of features for 4.13

New uabi:
- batch bo in first slot, for faster execbuf assembly in userspace
  (Chris Wilson)
- (sub)slice getparam, needed for mesa perf support (Robert Bragg)

First pile of patches for cnl/cfl support, maintained by Rodrigo but
with lots of contributions from others. Still incomplete since public
review still ongoing.

Features/refactoring:
- Make execbuf faster (Chris Wilson), a pile of series to make execbuf
  buffer handling have fewer passes, use less list walking, postpone
  more work to async workers and shuffle buffers less, all to make the
  common case much faster (in some cases at least).
- cold boot support for glk dsi (Madhav Chauhan)
- Clean up pipe A quirk and related old platform hacks (Ville)
- perf sampling support for kbl/glk (Lionel)
- perf cleanups (Robert Bragg)
- wire atomic state to backlight code, to avoid pipe lookup hacks
  (Maarten)
- reduce request waiting latency/overhead to remove the spinning and
  associated cpu cycle wasting (Chris)
- fix 90/270 rotation wm computation (Ville)
- new ddb allocation algo for skl (Kumar Mahesh)
- fix regression due to system suspend optimiazatino (Imre)
- the usual pile of small cleanups and refactors all over

GVT updates contained in this tag:
- optimization for per-VM mmio save/restore (Changbin)
- optimization for mmio hash table (Changbin)
- scheduler optimization with event (Ping)
- vGPU reset refinement (Fred)
- other misc refactor and cleanups, etc.

* tag 'drm-intel-next-2017-06-19' of git://anongit.freedesktop.org/git/drm-intel: (170 commits)
  drm/i915: Update DRIVER_DATE to 20170619
  drm/i915/cfl: Introduce Coffee Lake workarounds.
  drm/i915: Store 9 bits of PCI Device ID for platforms with a LP PCH
  drm/i915: Stash a pointer to the obj's resv in the vma
  drm/i915: Async GPU relocation processing
  drm/i915: Allow execbuffer to use the first object as the batch
  drm/i915: Wait upon userptr get-user-pages within execbuffer
  drm/i915: First try the previous execbuffer location
  drm/i915: Store a persistent reference for an object in the execbuffer cache
  drm/i915: Eliminate lots of iterations over the execobjects array
  drm/i915: Disable EXEC_OBJECT_ASYNC when doing relocations
  drm/i915: Pass vma to relocate entry
  drm/i915: Store a direct lookup from object handle to vma
  drm/i915: Fix retrieval of hangcheck stats
  drm/i915: Store i915_gem_object_is_coherent() as a bit next to cache-dirty
  drm/i915: Mark CPU cache as dirty on every transition for CPU writes
  drm/i915: Make i915_vma_destroy() static
  drm/i915: Actually attach the tv_format property to the SDVO connector
  Revert "drm/i915/skl: New ddb allocation algorithm"
  drm/i915/glk: Add cold boot sequence for GLK DSI
  ...

108 files changed, 36835 insertions, 3297 deletions

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 16dccf550412..f8227318dcaf 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -129,7 +129,16 @@ i915-y += i915_vgpu.o
 
 # perf code
 i915-y += i915_perf.o \
-	  i915_oa_hsw.o
+	  i915_oa_hsw.o \
+	  i915_oa_bdw.o \
+	  i915_oa_chv.o \
+	  i915_oa_sklgt2.o \
+	  i915_oa_sklgt3.o \
+	  i915_oa_sklgt4.o \
+	  i915_oa_bxt.o \
+	  i915_oa_kblgt2.o \
+	  i915_oa_kblgt3.o \
+	  i915_oa_glk.o
 
 ifeq ($(CONFIG_DRM_I915_GVT),y)
 i915-y += intel_gvt.o
diff --git a/drivers/gpu/drm/i915/dvo_ch7xxx.c b/drivers/gpu/drm/i915/dvo_ch7xxx.c
index 44b3159f2fe8..7aeeffd2428b 100644
--- a/drivers/gpu/drm/i915/dvo_ch7xxx.c
+++ b/drivers/gpu/drm/i915/dvo_ch7xxx.c
@@ -217,9 +217,8 @@ static bool ch7xxx_init(struct intel_dvo_device *dvo,
 
 	name = ch7xxx_get_id(vendor);
 	if (!name) {
-		DRM_DEBUG_KMS("ch7xxx not detected; got 0x%02x from %s "
-				"slave %d.\n",
-			  vendor, adapter->name, dvo->slave_addr);
+		DRM_DEBUG_KMS("ch7xxx not detected; got VID 0x%02x from %s slave %d.\n",
+			      vendor, adapter->name, dvo->slave_addr);
 		goto out;
 	}
 
@@ -229,9 +228,8 @@ static bool ch7xxx_init(struct intel_dvo_device *dvo,
 
 	devid = ch7xxx_get_did(device);
 	if (!devid) {
-		DRM_DEBUG_KMS("ch7xxx not detected; got 0x%02x from %s "
-				"slave %d.\n",
-			  vendor, adapter->name, dvo->slave_addr);
+		DRM_DEBUG_KMS("ch7xxx not detected; got DID 0x%02x from %s slave %d.\n",
+			      device, adapter->name, dvo->slave_addr);
 		goto out;
 	}
 
diff --git a/drivers/gpu/drm/i915/gvt/Makefile b/drivers/gpu/drm/i915/gvt/Makefile
index b123c20e2097..f5486cb94818 100644
--- a/drivers/gpu/drm/i915/gvt/Makefile
+++ b/drivers/gpu/drm/i915/gvt/Makefile
@@ -3,6 +3,6 @@ GVT_SOURCE := gvt.o aperture_gm.o handlers.o vgpu.o trace_points.o firmware.o \
 	interrupt.o gtt.o cfg_space.o opregion.o mmio.o display.o edid.o \
 	execlist.o scheduler.o sched_policy.o render.o cmd_parser.o
 
-ccflags-y				+= -I$(src) -I$(src)/$(GVT_DIR) -Wall
+ccflags-y				+= -I$(src) -I$(src)/$(GVT_DIR)
 i915-y					+= $(addprefix $(GVT_DIR)/, $(GVT_SOURCE))
 obj-$(CONFIG_DRM_I915_GVT_KVMGT)	+= $(GVT_DIR)/kvmgt.o
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 41b2c3aaa04a..51241de5e7a7 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -2414,53 +2414,13 @@ static void add_cmd_entry(struct intel_gvt *gvt, struct cmd_entry *e)
 	hash_add(gvt->cmd_table, &e->hlist, e->info->opcode);
 }
 
-#define GVT_MAX_CMD_LENGTH     20  /* In Dword */
-
-static void trace_cs_command(struct parser_exec_state *s,
-		cycles_t cost_pre_cmd_handler, cycles_t cost_cmd_handler)
-{
-	/* This buffer is used by ftrace to store all commands copied from
-	 * guest gma space. Sometimes commands can cross pages, this should
-	 * not be handled in ftrace logic. So this is just used as a
-	 * 'bounce buffer'
-	 */
-	u32 cmd_trace_buf[GVT_MAX_CMD_LENGTH];
-	int i;
-	u32 cmd_len = cmd_length(s);
-	/* The chosen value of GVT_MAX_CMD_LENGTH are just based on
-	 * following two considerations:
-	 * 1) From observation, most common ring commands is not that long.
-	 *    But there are execeptions. So it indeed makes sence to observe
-	 *    longer commands.
-	 * 2) From the performance and debugging point of view, dumping all
-	 *    contents of very commands is not necessary.
-	 * We mgith shrink GVT_MAX_CMD_LENGTH or remove this trace event in
-	 * future for performance considerations.
-	 */
-	if (unlikely(cmd_len > GVT_MAX_CMD_LENGTH)) {
-		gvt_dbg_cmd("cmd length exceed tracing limitation!\n");
-		cmd_len = GVT_MAX_CMD_LENGTH;
-	}
-
-	for (i = 0; i < cmd_len; i++)
-		cmd_trace_buf[i] = cmd_val(s, i);
-
-	trace_gvt_command(s->vgpu->id, s->ring_id, s->ip_gma, cmd_trace_buf,
-			cmd_len, s->buf_type == RING_BUFFER_INSTRUCTION,
-			cost_pre_cmd_handler, cost_cmd_handler);
-}
-
 /* call the cmd handler, and advance ip */
 static int cmd_parser_exec(struct parser_exec_state *s)
 {
+	struct intel_vgpu *vgpu = s->vgpu;
 	struct cmd_info *info;
 	u32 cmd;
 	int ret = 0;
-	cycles_t t0, t1, t2;
-	struct parser_exec_state s_before_advance_custom;
-	struct intel_vgpu *vgpu = s->vgpu;
-
-	t0 = get_cycles();
 
 	cmd = cmd_val(s, 0);
 
@@ -2471,13 +2431,10 @@ static int cmd_parser_exec(struct parser_exec_state *s)
 		return -EINVAL;
 	}
 
-	gvt_dbg_cmd("%s\n", info->name);
-
 	s->info = info;
 
-	t1 = get_cycles();
-
-	s_before_advance_custom = *s;
+	trace_gvt_command(vgpu->id, s->ring_id, s->ip_gma, s->ip_va,
+			  cmd_length(s), s->buf_type);
 
 	if (info->handler) {
 		ret = info->handler(s);
@@ -2486,9 +2443,6 @@ static int cmd_parser_exec(struct parser_exec_state *s)
 			return ret;
 		}
 	}
-	t2 = get_cycles();
-
-	trace_cs_command(&s_before_advance_custom, t1 - t0, t2 - t1);
 
 	if (!(info->flag & F_IP_ADVANCE_CUSTOM)) {
 		ret = cmd_advance_default(s);