Refined set of intel_ringbuffer renames

Refined set of intel_ringbuffer renames

[Chris Wilson]

In preparation for having to only apply bug fixes to the code once, we
first need to unify request construction. Joonas has reviewed almost
all of the patches, and then added some more...
-Chris

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()

[Chris Wilson]

Since intel_engine_enable_signaling() is now only called via
fence_enable_sw_signaling(), we can rely on it to provide serialisation
and run-once for us and so make ourselves slightly simpler.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_breadcrumbs.c | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c
index d893ccdd62ac..90867446f1a5 100644
--- a/drivers/gpu/drm/i915/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c
@@ -480,19 +480,15 @@ void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
 	struct rb_node *parent, **p;
 	bool first, wakeup;
 
-	if (unlikely(READ_ONCE(request->signaling.wait.tsk)))
-		return;
-
-	spin_lock(&b->lock);
-	if (unlikely(request->signaling.wait.tsk)) {
-		wakeup = false;
-		goto unlock;
-	}
+	/* locked by fence_enable_sw_signaling() */
+	assert_spin_locked(&request->lock);
 
 	request->signaling.wait.tsk = b->signaler;
 	request->signaling.wait.seqno = request->fence.seqno;
 	i915_gem_request_get(request);
 
+	spin_lock(&b->lock);
+
 	/* First add ourselves into the list of waiters, but register our
 	 * bottom-half as the signaller thread. As per usual, only the oldest
 	 * waiter (not just signaller) is tasked as the bottom-half waking
@@ -525,7 +521,6 @@ void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
 	if (first)
 		smp_store_mb(b->first_signal, request);
 
-unlock:
 	spin_unlock(&b->lock);
 
 	if (wakeup)
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 02/31] drm/i915: Prefer list_first_entry_or_null

[Chris Wilson]

list_first_entry_or_null() can generate better code than using
if (!list_empty()) {ptr = list_first_entry()) ..., so put it to use.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c      | 12 +++++-------
 drivers/gpu/drm/i915/i915_gem_request.c  |  8 +++-----
 drivers/gpu/drm/i915/i915_gem_shrinker.c |  9 +++++----
 3 files changed, 13 insertions(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 30da543e1bdf..38e7d992a20d 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -2736,13 +2736,11 @@ static void i915_gtt_color_adjust(struct drm_mm_node *node,
 	if (node->color != color)
 		*start += 4096;
 
-	if (!list_empty(&node->node_list)) {
-		node = list_entry(node->node_list.next,
-				  struct drm_mm_node,
-				  node_list);
-		if (node->allocated && node->color != color)
-			*end -= 4096;
-	}
+	node = list_first_entry_or_null(&node->node_list,
+					struct drm_mm_node,
+					node_list);
+	if (node && node->allocated && node->color != color)
+		*end -= 4096;
 }
 
 static int i915_gem_setup_global_gtt(struct drm_device *dev,
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 60a3a343b3a8..49396b895a36 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -317,12 +317,10 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
 		return ret;
 
 	/* Move the oldest request to the slab-cache (if not in use!) */
-	if (!list_empty(&engine->request_list)) {
-		req = list_first_entry(&engine->request_list,
+	req = list_first_entry_or_null(&engine->request_list,
 				       typeof(*req), list);
-		if (i915_gem_request_completed(req))
-			i915_gem_request_retire(req);
-	}
+	if (req && i915_gem_request_completed(req))
+		i915_gem_request_retire(req);
 
 	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
 	if (!req)
diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
index afaa2597e35e..5d4772c146b1 100644
--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
@@ -163,17 +163,18 @@ i915_gem_shrink(struct drm_i915_private *dev_priv,
 	 */
 	for (phase = phases; phase->list; phase++) {
 		struct list_head still_in_list;
+		struct drm_i915_gem_object *obj;
 
 		if ((flags & phase->bit) == 0)
 			continue;
 
 		INIT_LIST_HEAD(&still_in_list);
-		while (count < target && !list_empty(phase->list)) {
-			struct drm_i915_gem_object *obj;
+		while (count < target &&
+		       (obj = list_first_entry_or_null(phase->list,
+						       typeof(*obj),
+						       global_list))) {
 			struct i915_vma *vma, *v;
 
-			obj = list_first_entry(phase->list,
-					       typeof(*obj), global_list);
 			list_move_tail(&obj->global_list, &still_in_list);
 
 			if (flags & I915_SHRINK_PURGEABLE &&
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 03/31] drm/i915: Only clear the client pointer when tearing down the file

[Chris Wilson]

Upon release of the file (i.e. the user calls close(fd)), we decouple
all objects from the client list so that we don't chase the dangling
file_priv. As we always inspect file_priv first, we only need to nullify
that pointer and can safely ignore the list_head.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index c8436639b3ed..3730aecc1eae 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4745,21 +4745,15 @@ int i915_gem_freeze_late(struct drm_i915_private *dev_priv)
 void i915_gem_release(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
+	struct drm_i915_gem_request *request;
 
 	/* Clean up our request list when the client is going away, so that
 	 * later retire_requests won't dereference our soon-to-be-gone
 	 * file_priv.
 	 */
 	spin_lock(&file_priv->mm.lock);
-	while (!list_empty(&file_priv->mm.request_list)) {
-		struct drm_i915_gem_request *request;
-
-		request = list_first_entry(&file_priv->mm.request_list,
-					   struct drm_i915_gem_request,
-					   client_list);
-		list_del(&request->client_list);
+	list_for_each_entry(request, &file_priv->mm.request_list, client_list)
 		request->file_priv = NULL;
-	}
 	spin_unlock(&file_priv->mm.lock);
 
 	if (!list_empty(&file_priv->rps.link)) {
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 04/31] drm/i915: Only drop the batch-pool's object reference

[Chris Wilson]

The obj->batch_pool_link is only inspected when traversing the batch
pool list and when on the batch pool list the object is referenced. Thus
when freeing the batch pool list, we only need to unreference the object
and do not have to worry about the obj->batch_pool_link.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_batch_pool.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
index 3507b2753fd3..825981b5aa40 100644
--- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c
+++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
@@ -68,15 +68,14 @@ void i915_gem_batch_pool_fini(struct i915_gem_batch_pool *pool)
 	WARN_ON(!mutex_is_locked(&pool->dev->struct_mutex));
 
 	for (n = 0; n < ARRAY_SIZE(pool->cache_list); n++) {
-		while (!list_empty(&pool->cache_list[n])) {
-			struct drm_i915_gem_object *obj =
-				list_first_entry(&pool->cache_list[n],
-						 struct drm_i915_gem_object,
-						 batch_pool_link);
+		struct drm_i915_gem_object *obj, *next;
 
-			list_del(&obj->batch_pool_link);
+		list_for_each_entry_safe(obj, next,
+					 &pool->cache_list[n],
+					 batch_pool_link)
 			i915_gem_object_put(obj);
-		}
+
+		INIT_LIST_HEAD(&pool->cache_list[n]);
 	}
 }
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 05/31] drm/i915/cmdparser: Remove stray intel_engine_cs *ring

[Chris Wilson]

When we refer to intel_engine_cs, we want to use engine so as not to
confuse ourselves about ringbuffers.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_cmd_parser.c  |  4 ++--
 drivers/gpu/drm/i915/i915_drv.h         |  5 +++--
 drivers/gpu/drm/i915/intel_ringbuffer.h | 10 +++++-----
 3 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index b0fd6a7b0603..b5b520176c59 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -636,7 +636,7 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
 	return ret;
 }
 
-static bool check_sorted(int ring_id,
+static bool check_sorted(int engine_id,
 			 const struct drm_i915_reg_descriptor *reg_table,
 			 int reg_count)
 {
@@ -649,7 +649,7 @@ static bool check_sorted(int ring_id,
 
 		if (curr < previous) {
 			DRM_ERROR("CMD: table not sorted ring=%d entry=%d reg=0x%08X prev=0x%08X\n",
-				  ring_id, i, curr, previous);
+				  engine_id, i, curr, previous);
 			ret = false;
 		}
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 9f655e2add66..d4c5c2edf252 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2500,8 +2500,9 @@ struct drm_i915_cmd_descriptor {
 /*
  * A table of commands requiring special handling by the command parser.
  *
- * Each ring has an array of tables. Each table consists of an array of command
- * descriptors, which must be sorted with command opcodes in ascending order.
+ * Each engine has an array of tables. Each table consists of an array of
+ * command descriptors, which must be sorted with command opcodes in
+ * ascending order.
  */
 struct drm_i915_cmd_table {
 	const struct drm_i915_cmd_descriptor *table;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 0f8019488d33..0a830c116886 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -340,7 +340,7 @@ struct intel_engine_cs {
 
 	/*
 	 * Table of commands the command parser needs to know about
-	 * for this ring.
+	 * for this engine.
 	 */
 	DECLARE_HASHTABLE(cmd_hash, I915_CMD_HASH_ORDER);
 
@@ -354,11 +354,11 @@ struct intel_engine_cs {
 	 * Returns the bitmask for the length field of the specified command.
 	 * Return 0 for an unrecognized/invalid command.
 	 *
-	 * If the command parser finds an entry for a command in the ring's
+	 * If the command parser finds an entry for a command in the engines's
 	 * cmd_tables, it gets the command's length based on the table entry.
-	 * If not, it calls this function to determine the per-ring length field
-	 * encoding for the command (i.e. certain opcode ranges use certain bits
-	 * to encode the command length in the header).
+	 * If not, it calls this function to determine the per-engine length
+	 * field encoding for the command (i.e. different opcode ranges use
+	 * certain bits to encode the command length in the header).
 	 */
 	u32 (*get_cmd_length_mask)(u32 cmd_header);
 };
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 06/31] drm/i915: Use engine to refer to the user's BSD intel_engine_cs

[Chris Wilson]

This patch transitions the execbuf engine selection away from using the
ring nomenclature - though we still refer to the user's incoming
selector as their user_ring_id.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_drv.h            |  4 ++--
 drivers/gpu/drm/i915/i915_gem.c            |  2 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 15 ++++++++-------
 3 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index d4c5c2edf252..b686b741ea87 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -402,7 +402,7 @@ struct drm_i915_file_private {
 		unsigned boosts;
 	} rps;
 
-	unsigned int bsd_ring;
+	unsigned int bsd_engine;
 };
 
 /* Used by dp and fdi links */
@@ -1331,7 +1331,7 @@ struct i915_gem_mm {
 	bool interruptible;
 
 	/* the indicator for dispatch video commands on two BSD rings */
-	unsigned int bsd_ring_dispatch_index;
+	unsigned int bsd_engine_dispatch_index;
 
 	/** Bit 6 swizzling required for X tiling */
 	uint32_t bit_6_swizzle_x;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 3730aecc1eae..e155e8dd28ed 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4782,7 +4782,7 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file)
 	spin_lock_init(&file_priv->mm.lock);
 	INIT_LIST_HEAD(&file_priv->mm.request_list);
 
-	file_priv->bsd_ring = -1;
+	file_priv->bsd_engine = -1;
 
 	ret = i915_gem_context_open(dev, file);
 	if (ret)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index f8d8ae39fc2f..9ef68e86b483 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1351,23 +1351,24 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 
 /**
  * Find one BSD ring to dispatch the corresponding BSD command.
- * The ring index is returned.
+ * The engine index is returned.
  */
 static unsigned int
-gen8_dispatch_bsd_ring(struct drm_i915_private *dev_priv, struct drm_file *file)
+gen8_dispatch_bsd_engine(struct drm_i915_private *dev_priv,
+			 struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 
 	/* Check whether the file_priv has already selected one ring. */
-	if ((int)file_priv->bsd_ring < 0) {
+	if ((int)file_priv->bsd_engine < 0) {
 		/* If not, use the ping-pong mechanism to select one. */
 		mutex_lock(&dev_priv->drm.struct_mutex);
-		file_priv->bsd_ring = dev_priv->mm.bsd_ring_dispatch_index;
-		dev_priv->mm.bsd_ring_dispatch_index ^= 1;
+		file_priv->bsd_engine = dev_priv->mm.bsd_engine_dispatch_index;
+		dev_priv->mm.bsd_engine_dispatch_index ^= 1;
 		mutex_unlock(&dev_priv->drm.struct_mutex);
 	}
 
-	return file_priv->bsd_ring;
+	return file_priv->bsd_engine;
 }
 
 #define I915_USER_RINGS (4)
@@ -1404,7 +1405,7 @@ eb_select_engine(struct drm_i915_private *dev_priv,
 		unsigned int bsd_idx = args->flags & I915_EXEC_BSD_MASK;
 
 		if (bsd_idx == I915_EXEC_BSD_DEFAULT) {
-			bsd_idx = gen8_dispatch_bsd_ring(dev_priv, file);
+			bsd_idx = gen8_dispatch_bsd_engine(dev_priv, file);
 		} else if (bsd_idx >= I915_EXEC_BSD_RING1 &&
 			   bsd_idx <= I915_EXEC_BSD_RING2) {
 			bsd_idx >>= I915_EXEC_BSD_SHIFT;
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 07/31] drm/i915: Avoid using intel_engine_cs *ring for GPU error capture

[Chris Wilson]

Inside the error capture itself, we refer to not only the hardware
engine, its ringbuffer but also the capture state. Finding clear names
for each whilst avoiding mixing ring/intel_engine_cs is tricky. As a
compromise we keep using ering for the error capture.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_drv.h       |   6 +-
 drivers/gpu/drm/i915/i915_gpu_error.c | 255 +++++++++++++++++-----------------
 2 files changed, 134 insertions(+), 127 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index b686b741ea87..7e24ea1d465c 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -512,7 +512,7 @@ struct drm_i915_error_state {
 	struct intel_display_error_state *display;
 	struct drm_i915_error_object *semaphore_obj;
 
-	struct drm_i915_error_ring {
+	struct drm_i915_error_engine {
 		bool valid;
 		/* Software tracked state */
 		bool waiting;
@@ -578,7 +578,7 @@ struct drm_i915_error_state {
 
 		pid_t pid;
 		char comm[TASK_COMM_LEN];
-	} ring[I915_NUM_ENGINES];
+	} engine[I915_NUM_ENGINES];
 
 	struct drm_i915_error_buffer {
 		u32 size;
@@ -593,7 +593,7 @@ struct drm_i915_error_state {
 		u32 dirty:1;
 		u32 purgeable:1;
 		u32 userptr:1;
-		s32 ring:4;
+		s32 engine:4;
 		u32 cache_level:3;
 	} **active_bo, **pinned_bo;
 
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 4d39c7284605..0ec917fa115d 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -30,9 +30,9 @@
 #include <generated/utsrelease.h>
 #include "i915_drv.h"
 
-static const char *ring_str(int ring)
+static const char *engine_str(int engine)
 {
-	switch (ring) {
+	switch (engine) {
 	case RCS: return "render";
 	case VCS: return "bsd";
 	case BCS: return "blt";
@@ -207,8 +207,8 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 		err_puts(m, dirty_flag(err->dirty));
 		err_puts(m, purgeable_flag(err->purgeable));
 		err_puts(m, err->userptr ? " userptr" : "");
-		err_puts(m, err->ring != -1 ? " " : "");
-		err_puts(m, ring_str(err->ring));
+		err_puts(m, err->engine != -1 ? " " : "");
+		err_puts(m, engine_str(err->engine));
 		err_puts(m, i915_cache_level_str(m->i915, err->cache_level));
 
 		if (err->name)
@@ -240,69 +240,71 @@ static const char *hangcheck_action_to_str(enum intel_ring_hangcheck_action a)
 }
 
 static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
-				  struct drm_device *dev,
-				  struct drm_i915_error_state *error,
-				  int ring_idx)
+				    struct drm_device *dev,
+				    struct drm_i915_error_state *error,
+				    int engine_idx)
 {
-	struct drm_i915_error_ring *ring = &error->ring[ring_idx];
+	struct drm_i915_error_engine *ering = &error->engine[engine_idx];
 
-	if (!ring->valid)
+	if (!ering->valid)
 		return;
 
-	err_printf(m, "%s command stream:\n", ring_str(ring_idx));
-	err_printf(m, "  START: 0x%08x\n", ring->start);
-	err_printf(m, "  HEAD:  0x%08x\n", ring->head);
-	err_printf(m, "  TAIL:  0x%08x\n", ring->tail);
-	err_printf(m, "  CTL:   0x%08x\n", ring->ctl);
-	err_printf(m, "  HWS:   0x%08x\n", ring->hws);
-	err_printf(m, "  ACTHD: 0x%08x %08x\n", (u32)(ring->acthd>>32), (u32)ring->acthd);
-	err_printf(m, "  IPEIR: 0x%08x\n", ring->ipeir);
-	err_printf(m, "  IPEHR: 0x%08x\n", ring->ipehr);
-	err_printf(m, "  INSTDONE: 0x%08x\n", ring->instdone);
+	err_printf(m, "%s command stream:\n", engine_str(engine_idx));
+	err_printf(m, "  START: 0x%08x\n", ering->start);
+	err_printf(m, "  HEAD:  0x%08x\n", ering->head);
+	err_printf(m, "  TAIL:  0x%08x\n", ering->tail);
+	err_printf(m, "  CTL:   0x%08x\n", ering->ctl);
+	err_printf(m, "  HWS:   0x%08x\n", ering->hws);
+	err_printf(m, "  ACTHD: 0x%08x %08x\n",
+		   (u32)(ering->acthd>>32), (u32)ering->acthd);
+	err_printf(m, "  IPEIR: 0x%08x\n", ering->ipeir);
+	err_printf(m, "  IPEHR: 0x%08x\n", ering->ipehr);
+	err_printf(m, "  INSTDONE: 0x%08x\n", ering->instdone);
 	if (INTEL_INFO(dev)->gen >= 4) {
-		err_printf(m, "  BBADDR: 0x%08x %08x\n", (u32)(ring->bbaddr>>32), (u32)ring->bbaddr);
-		err_printf(m, "  BB_STATE: 0x%08x\n", ring->bbstate);
-		err_printf(m, "  INSTPS: 0x%08x\n", ring->instps);
+		err_printf(m, "  BBADDR: 0x%08x %08x\n",
+			   (u32)(ering->bbaddr>>32), (u32)ering->bbaddr);
+		err_printf(m, "  BB_STATE: 0x%08x\n", ering->bbstate);
+		err_printf(m, "  INSTPS: 0x%08x\n", ering->instps);
 	}
-	err_printf(m, "  INSTPM: 0x%08x\n", ring->instpm);
-	err_printf(m, "  FADDR: 0x%08x %08x\n", upper_32_bits(ring->faddr),
-		   lower_32_bits(ring->faddr));
+	err_printf(m, "  INSTPM: 0x%08x\n", ering->instpm);
+	err_printf(m, "  FADDR: 0x%08x %08x\n", upper_32_bits(ering->faddr),
+		   lower_32_bits(ering->faddr));
 	if (INTEL_INFO(dev)->gen >= 6) {
-		err_printf(m, "  RC PSMI: 0x%08x\n", ring->rc_psmi);
-		err_printf(m, "  FAULT_REG: 0x%08x\n", ring->fault_reg);
+		err_printf(m, "  RC PSMI: 0x%08x\n", ering->rc_psmi);
+		err_printf(m, "  FAULT_REG: 0x%08x\n", ering->fault_reg);
 		err_printf(m, "  SYNC_0: 0x%08x [last synced 0x%08x]\n",
-			   ring->semaphore_mboxes[0],
-			   ring->semaphore_seqno[0]);
+			   ering->semaphore_mboxes[0],
+			   ering->semaphore_seqno[0]);
 		err_printf(m, "  SYNC_1: 0x%08x [last synced 0x%08x]\n",
-			   ring->semaphore_mboxes[1],
-			   ring->semaphore_seqno[1]);
+			   ering->semaphore_mboxes[1],
+			   ering->semaphore_seqno[1]);
 		if (HAS_VEBOX(dev)) {
 			err_printf(m, "  SYNC_2: 0x%08x [last synced 0x%08x]\n",
-				   ring->semaphore_mboxes[2],
-				   ring->semaphore_seqno[2]);
+				   ering->semaphore_mboxes[2],
+				   ering->semaphore_seqno[2]);
 		}
 	}
 	if (USES_PPGTT(dev)) {
-		err_printf(m, "  GFX_MODE: 0x%08x\n", ring->vm_info.gfx_mode);
+		err_printf(m, "  GFX_MODE: 0x%08x\n", ering->vm_info.gfx_mode);
 
 		if (INTEL_INFO(dev)->gen >= 8) {
 			int i;
 			for (i = 0; i < 4; i++)
 				err_printf(m, "  PDP%d: 0x%016llx\n",
-					   i, ring->vm_info.pdp[i]);
+					   i, ering->vm_info.pdp[i]);
 		} else {
 			err_printf(m, "  PP_DIR_BASE: 0x%08x\n",
-				   ring->vm_info.pp_dir_base);
+				   ering->vm_info.pp_dir_base);
 		}
 	}
-	err_printf(m, "  seqno: 0x%08x\n", ring->seqno);
-	err_printf(m, "  last_seqno: 0x%08x\n", ring->last_seqno);
-	err_printf(m, "  waiting: %s\n", yesno(ring->waiting));
-	err_printf(m, "  ring->head: 0x%08x\n", ring->cpu_ring_head);
-	err_printf(m, "  ring->tail: 0x%08x\n", ring->cpu_ring_tail);
+	err_printf(m, "  seqno: 0x%08x\n", ering->seqno);
+	err_printf(m, "  last_seqno: 0x%08x\n", ering->last_seqno);
+	err_printf(m, "  waiting: %s\n", yesno(ering->waiting));
+	err_printf(m, "  ring->head: 0x%08x\n", ering->cpu_ring_head);
+	err_printf(m, "  ring->tail: 0x%08x\n", ering->cpu_ring_tail);
 	err_printf(m, "  hangcheck: %s [%d]\n",
-		   hangcheck_action_to_str(ring->hangcheck_action),
-		   ring->hangcheck_score);
+		   hangcheck_action_to_str(ering->hangcheck_action),
+		   ering->hangcheck_score);
 }
 
 void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...)
@@ -348,17 +350,17 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 		   error->time.tv_usec);
 	err_printf(m, "Kernel: " UTS_RELEASE "\n");
 	max_hangcheck_score = 0;
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
-		if (error->ring[i].hangcheck_score > max_hangcheck_score)
-			max_hangcheck_score = error->ring[i].hangcheck_score;
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		if (error->engine[i].hangcheck_score > max_hangcheck_score)
+			max_hangcheck_score = error->engine[i].hangcheck_score;
 	}
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
-		if (error->ring[i].hangcheck_score == max_hangcheck_score &&
-		    error->ring[i].pid != -1) {
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		if (error->engine[i].hangcheck_score == max_hangcheck_score &&
+		    error->engine[i].pid != -1) {
 			err_printf(m, "Active process (on ring %s): %s [%d]\n",
-				   ring_str(i),
-				   error->ring[i].comm,
-				   error->ring[i].pid);
+				   engine_str(i),
+				   error->engine[i].comm,
+				   error->engine[i].pid);
 		}
 	}
 	err_printf(m, "Reset count: %u\n", error->reset_count);
@@ -414,7 +416,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	if (IS_GEN7(dev))
 		err_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
 
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++)
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++)
 		i915_ring_error_state(m, dev, error, i);
 
 	for (i = 0; i < error->vm_count; i++) {
@@ -429,21 +431,23 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 				    error->pinned_bo_count[i]);
 	}
 
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
-		obj = error->ring[i].batchbuffer;
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		struct drm_i915_error_engine *ering = &error->engine[i];
+
+		obj = ering->batchbuffer;
 		if (obj) {
 			err_puts(m, dev_priv->engine[i].name);
-			if (error->ring[i].pid != -1)
+			if (ering->pid != -1)
 				err_printf(m, " (submitted by %s [%d])",
-					   error->ring[i].comm,
-					   error->ring[i].pid);
+					   ering->comm,
+					   ering->pid);
 			err_printf(m, " --- gtt_offset = 0x%08x %08x\n",
 				   upper_32_bits(obj->gtt_offset),
 				   lower_32_bits(obj->gtt_offset));
 			print_error_obj(m, obj);
 		}
 
-		obj = error->ring[i].wa_batchbuffer;
+		obj = ering->wa_batchbuffer;
 		if (obj) {
 			err_printf(m, "%s (w/a) --- gtt_offset = 0x%08x\n",
 				   dev_priv->engine[i].name,
@@ -451,38 +455,38 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			print_error_obj(m, obj);
 		}
 
-		if (error->ring[i].num_requests) {
+		if (ering->num_requests) {
 			err_printf(m, "%s --- %d requests\n",
 				   dev_priv->engine[i].name,
-				   error->ring[i].num_requests);
-			for (j = 0; j < error->ring[i].num_requests; j++) {
+				   ering->num_requests);
+			for (j = 0; j < ering->num_requests; j++) {
 				err_printf(m, "  seqno 0x%08x, emitted %ld, tail 0x%08x\n",
-					   error->ring[i].requests[j].seqno,
-					   error->ring[i].requests[j].jiffies,
-					   error->ring[i].requests[j].tail);
+					   ering->requests[j].seqno,
+					   ering->requests[j].jiffies,
+					   ering->requests[j].tail);
 			}
 		}
 
-		if (error->ring[i].num_waiters) {
+		if (ering->num_waiters) {
 			err_printf(m, "%s --- %d waiters\n",
 				   dev_priv->engine[i].name,
-				   error->ring[i].num_waiters);
-			for (j = 0; j < error->ring[i].num_waiters; j++) {
+				   ering->num_waiters);
+			for (j = 0; j < ering->num_waiters; j++) {
 				err_printf(m, " seqno 0x%08x for %s [%d]\n",
-					   error->ring[i].waiters[j].seqno,
-					   error->ring[i].waiters[j].comm,
-					   error->ring[i].waiters[j].pid);
+					   ering->waiters[j].seqno,
+					   ering->waiters[j].comm,
+					   ering->waiters[j].pid);
 			}
 		}
 
-		if ((obj = error->ring[i].ringbuffer)) {
+		if ((obj = ering->ringbuffer)) {
 			err_printf(m, "%s --- ringbuffer = 0x%08x\n",
 				   dev_priv->engine[i].name,
 				   lower_32_bits(obj->gtt_offset));
 			print_error_obj(m, obj);
 		}
 
-		if ((obj = error->ring[i].hws_page)) {
+		if ((obj = ering->hws_page)) {
 			u64 hws_offset = obj->gtt_offset;
 			u32 *hws_page = &obj->pages[0][0];
 
@@ -504,7 +508,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			}
 		}
 
-		obj = error->ring[i].wa_ctx;
+		obj = ering->wa_ctx;
 		if (obj) {
 			u64 wa_ctx_offset = obj->gtt_offset;
 			u32 *wa_ctx_page = &obj->pages[0][0];
@@ -526,7 +530,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			}
 		}
 
-		if ((obj = error->ring[i].ctx)) {
+		if ((obj = ering->ctx)) {
 			err_printf(m, "%s --- HW Context = 0x%08x\n",
 				   dev_priv->engine[i].name,
 				   lower_32_bits(obj->gtt_offset));
@@ -611,15 +615,18 @@ static void i915_error_state_free(struct kref *error_ref)
 							  typeof(*error), ref);
 	int i;
 
-	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
-		i915_error_object_free(error->ring[i].batchbuffer);
-		i915_error_object_free(error->ring[i].wa_batchbuffer);
-		i915_error_object_free(error->ring[i].ringbuffer);
-		i915_error_object_free(error->ring[i].hws_page);
-		i915_error_object_free(error->ring[i].ctx);
-		i915_error_object_free(error->ring[i].wa_ctx);
-		kfree(error->ring[i].requests);
-		kfree(error->ring[i].waiters);
+	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+		struct drm_i915_error_engine *ering = &error->engine[i];
+
+		i915_error_object_free(ering->batchbuffer);
+		i915_error_object_free(ering->wa_batchbuffer);
+		i915_error_object_free(ering->ringbuffer);
+		i915_error_object_free(ering->hws_page);
+		i915_error_object_free(ering->ctx);
+		i915_error_object_free(ering->wa_ctx);
+
+		kfree(ering->requests);
+		kfree(ering->waiters);
 	}
 
 	i915_error_object_free(error->semaphore_obj);
@@ -762,8 +769,8 @@ static void capture_bo(struct drm_i915_error_buffer *err,
 	err->dirty = obj->dirty;
 	err->purgeable = obj->madv != I915_MADV_WILLNEED;
 	err->userptr = obj->userptr.mm != NULL;
-	err->ring = obj->last_write_req ?
-			i915_gem_request_get_engine(obj->last_write_req)->id : -1;
+	err->engine = obj->last_write_req ?
+		i915_gem_request_get_engine(obj->last_write_req)->id : -1;
 	err->cache_level = obj->cache_level;
 }
 
@@ -815,7 +822,7 @@ static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
  */
 static uint32_t i915_error_generate_code(struct drm_i915_private *dev_priv,
 					 struct drm_i915_error_state *error,
-					 int *ring_id)
+					 int *engine_id)
 {
 	uint32_t error_code = 0;
 	int i;
@@ -826,11 +833,11 @@ static uint32_t i915_error_generate_code(struct drm_i915_private *dev_priv,
 	 * strictly a client bug. Use instdone to differentiate those some.
 	 */
 	for (i = 0; i < I915_NUM_ENGINES; i++) {
-		if (error->ring[i].hangcheck_action == HANGCHECK_HUNG) {
-			if (ring_id)
-				*ring_id = i;
+		if (error->engine[i].hangcheck_action == HANGCHECK_HUNG) {
+			if (engine_id)
+				*engine_id = i;
 
-			return error->ring[i].ipehr ^ error->ring[i].instdone;
+			return error->engine[i].ipehr ^ error->engine[i].instdone;
 		}
 	}
 
@@ -858,7 +865,7 @@ static void i915_gem_record_fences(struct drm_i915_private *dev_priv,
 static void gen8_record_semaphore_state(struct drm_i915_private *dev_priv,
 					struct drm_i915_error_state *error,
 					struct intel_engine_cs *engine,
-					struct drm_i915_error_ring *ering)
+					struct drm_i915_error_engine *ering)
 {
 	struct intel_engine_cs *to;
 	enum intel_engine_id id;
@@ -891,7 +898,7 @@ static void gen8_record_semaphore_state(struct drm_i915_private *dev_priv,
 
 static void gen6_record_semaphore_state(struct drm_i915_private *dev_priv,
 					struct intel_engine_cs *engine,
-					struct drm_i915_error_ring *ering)
+					struct drm_i915_error_engine *ering)
 {
 	ering->semaphore_mboxes[0] = I915_READ(RING_SYNC_0(engine->mmio_base));
 	ering->semaphore_mboxes[1] = I915_READ(RING_SYNC_1(engine->mmio_base));
@@ -906,7 +913,7 @@ static void gen6_record_semaphore_state(struct drm_i915_private *dev_priv,
 }
 
 static void engine_record_waiters(struct intel_engine_cs *engine,
-				  struct drm_i915_error_ring *ering)
+				  struct drm_i915_error_engine *ering)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 	struct drm_i915_error_waiter *waiter;
@@ -950,7 +957,7 @@ static void engine_record_waiters(struct intel_engine_cs *engine,
 static void i915_record_ring_state(struct drm_i915_private *dev_priv,
 				   struct drm_i915_error_state *error,
 				   struct intel_engine_cs *engine,
-				   struct drm_i915_error_ring *ering)
+				   struct drm_i915_error_engine *ering)
 {
 	if (INTEL_GEN(dev_priv) >= 6) {
 		ering->rc_psmi = I915_READ(RING_PSMI_CTL(engine->mmio_base));
@@ -1048,7 +1055,7 @@ static void i915_record_ring_state(struct drm_i915_private *dev_priv,
 
 static void i915_gem_record_active_context(struct intel_engine_cs *engine,
 					   struct drm_i915_error_state *error,
-					   struct drm_i915_error_ring *ering)
+					   struct drm_i915_error_engine *ering)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	struct drm_i915_gem_object *obj;
@@ -1077,16 +1084,17 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 
 	for (i = 0; i < I915_NUM_ENGINES; i++) {
 		struct intel_engine_cs *engine = &dev_priv->engine[i];
+		struct drm_i915_error_engine *ering = &error->engine[i];
 
-		error->ring[i].pid = -1;
+		ering->pid = -1;
 
 		if (!intel_engine_initialized(engine))
 			continue;
 
-		error->ring[i].valid = true;
+		ering->valid = true;
 
-		i915_record_ring_state(dev_priv, error, engine, &error->ring[i]);
-		engine_record_waiters(engine, &error->ring[i]);
+		i915_record_ring_state(dev_priv, error, engine, ering);
+		engine_record_waiters(engine, ering);
 
 		request = i915_gem_find_active_request(engine);
 		if (request) {
@@ -1100,13 +1108,13 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 			 * as the simplest method to avoid being overwritten
 			 * by userspace.
 			 */
-			error->ring[i].batchbuffer =
+			ering->batchbuffer =
 				i915_error_object_create(dev_priv,
 							 request->batch_obj,
 							 vm);
 
 			if (HAS_BROKEN_CS_TLB(dev_priv))
-				error->ring[i].wa_batchbuffer =
+				ering->wa_batchbuffer =
 					i915_error_ggtt_object_create(dev_priv,
 							     engine->scratch.obj);
 
@@ -1116,8 +1124,8 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 				rcu_read_lock();
 				task = pid_task(request->pid, PIDTYPE_PID);
 				if (task) {
-					strcpy(error->ring[i].comm, task->comm);
-					error->ring[i].pid = task->pid;
+					strcpy(ering->comm, task->comm);
+					ering->pid = task->pid;
 				}
 				rcu_read_unlock();
 			}
@@ -1126,35 +1134,34 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 				request->ctx->flags & CONTEXT_NO_ERROR_CAPTURE;
 
 			rb = request->ringbuf;
-			error->ring[i].cpu_ring_head = rb->head;
-			error->ring[i].cpu_ring_tail = rb->tail;
-			error->ring[i].ringbuffer =
+			ering->cpu_ring_head = rb->head;
+			ering->cpu_ring_tail = rb->tail;
+			ering->ringbuffer =
 				i915_error_ggtt_object_create(dev_priv,
 							      rb->obj);
 		}
 
-		error->ring[i].hws_page =
+		ering->hws_page =
 			i915_error_ggtt_object_create(dev_priv,
 						      engine->status_page.obj);
 
 		if (engine->wa_ctx.obj) {
-			error->ring[i].wa_ctx =
+			ering->wa_ctx =
 				i915_error_ggtt_object_create(dev_priv,
 							      engine->wa_ctx.obj);
 		}
 
-		i915_gem_record_active_context(engine, error, &error->ring[i]);
+		i915_gem_record_active_context(engine, error, ering);
 
 		count = 0;
 		list_for_each_entry(request, &engine->request_list, list)
 			count++;
 
-		error->ring[i].num_requests = count;
-		error->ring[i].requests =
-			kcalloc(count, sizeof(*error->ring[i].requests),
-				GFP_ATOMIC);
-		if (error->ring[i].requests == NULL) {
-			error->ring[i].num_requests = 0;
+		ering->num_requests = count;
+		ering->requests =
+			kcalloc(count, sizeof(*ering->requests), GFP_ATOMIC);
+		if (!ering->requests) {
+			ering->num_requests = 0;
 			continue;
 		}
 
@@ -1162,7 +1169,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 		list_for_each_entry(request, &engine->request_list, list) {
 			struct drm_i915_error_request *erq;
 
-			if (count >= error->ring[i].num_requests) {
+			if (count >= ering->num_requests) {
 				/*
 				 * If the ring request list was changed in
 				 * between the point where the error request
@@ -1181,7 +1188,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 				break;
 			}
 
-			erq = &error->ring[i].requests[count++];
+			erq = &ering->requests[count++];
 			erq->seqno = request->fence.seqno;
 			erq->jiffies = request->emitted_jiffies;
 			erq->tail = request->postfix;
@@ -1352,20 +1359,20 @@ static void i915_error_capture_msg(struct drm_i915_private *dev_priv,
 				   const char *error_msg)
 {
 	u32 ecode;
-	int ring_id = -1, len;
+	int engine_id = -1, len;
 
-	ecode = i915_error_generate_code(dev_priv, error, &ring_id);
+	ecode = i915_error_generate_code(dev_priv, error, &engine_id);
 
 	len = scnprintf(error->error_msg, sizeof(error->error_msg),
 			"GPU HANG: ecode %d:%d:0x%08x",
-			INTEL_GEN(dev_priv), ring_id, ecode);
+			INTEL_GEN(dev_priv), engine_id, ecode);
 
-	if (ring_id != -1 && error->ring[ring_id].pid != -1)
+	if (engine_id != -1 && error->engine[engine_id].pid != -1)
 		len += scnprintf(error->error_msg + len,
 				 sizeof(error->error_msg) - len,
 				 ", in %s [%d]",
-				 error->ring[ring_id].comm,
-				 error->ring[ring_id].pid);
+				 error->engine[engine_id].comm,
+				 error->engine[engine_id].pid);
 
 	scnprintf(error->error_msg + len, sizeof(error->error_msg) - len,
 		  ", reason: %s, action: %s",
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 08/31] drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c

[Chris Wilson]

A few places we use ring when referring to the struct intel_engine_cs. An
anachronism we are pruning out.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index e155e8dd28ed..7bfce1d5c61b 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -46,7 +46,7 @@ static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *o
 static void
 i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
 static void
-i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring);
+i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int engine);
 
 static bool cpu_cache_is_coherent(struct drm_device *dev,
 				  enum i915_cache_level level)
@@ -1385,10 +1385,10 @@ static void
 i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
 			       struct drm_i915_gem_request *req)
 {
-	int ring = req->engine->id;
+	int idx = req->engine->id;
 
-	if (obj->last_read_req[ring] == req)
-		i915_gem_object_retire__read(obj, ring);
+	if (obj->last_read_req[idx] == req)
+		i915_gem_object_retire__read(obj, idx);
 	else if (obj->last_write_req == req)
 		i915_gem_object_retire__write(obj);
 
@@ -2381,20 +2381,20 @@ i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
 }
 
 static void
-i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
+i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)
 {
 	struct i915_vma *vma;
 
-	GEM_BUG_ON(obj->last_read_req[ring] == NULL);
-	GEM_BUG_ON(!(obj->active & (1 << ring)));
+	GEM_BUG_ON(obj->last_read_req[idx] == NULL);
+	GEM_BUG_ON(!(obj->active & (1 << idx)));
 
-	list_del_init(&obj->engine_list[ring]);
-	i915_gem_request_assign(&obj->last_read_req[ring], NULL);
+	list_del_init(&obj->engine_list[idx]);
+	i915_gem_request_assign(&obj->last_read_req[idx], NULL);
 
-	if (obj->last_write_req && obj->last_write_req->engine->id == ring)
+	if (obj->last_write_req && obj->last_write_req->engine->id == idx)
 		i915_gem_object_retire__write(obj);
 
-	obj->active &= ~(1 << ring);
+	obj->active &= ~(1 << idx);
 	if (obj->active)
 		return;
 
@@ -4599,7 +4599,7 @@ int i915_gem_init(struct drm_device *dev)
 
 	ret = i915_gem_init_hw(dev);
 	if (ret == -EIO) {
-		/* Allow ring initialisation to fail by marking the GPU as
+		/* Allow engine initialisation to fail by marking the GPU as
 		 * wedged. But we only want to do this where the GPU is angry,
 		 * for all other failure, such as an allocation failure, bail.
 		 */
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 09/31] drm/i915: Update a couple of hangcheck comments to talk about engines

[Chris Wilson]

We still have lots of comments that refer to the old ring when we mean
struct intel_engine_cs and its hardware correspondence. This patch fixes
an instance inside hangcheck to talk about engines.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_irq.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 7104dc1463eb..f5bf4f913a91 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -3140,13 +3140,13 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 				}
 			} else {
 				/* We always increment the hangcheck score
-				 * if the ring is busy and still processing
+				 * if the engine is busy and still processing
 				 * the same request, so that no single request
 				 * can run indefinitely (such as a chain of
 				 * batches). The only time we do not increment
 				 * the hangcheck score on this ring, if this
-				 * ring is in a legitimate wait for another
-				 * ring. In that case the waiting ring is a
+				 * engine is in a legitimate wait for another
+				 * engine. In that case the waiting engine is a
 				 * victim and we want to be sure we catch the
 				 * right culprit. Then every time we do kick
 				 * the ring, add a small increment to the
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 10/31] drm/i915: Unify intel_logical_ring_emit and intel_ring_emit

[Chris Wilson]

Both perform the same actions with more or less indirection, so just
unify the code.

v2: Add back a few intel_engine_cs locals

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_context.c    |  47 ++--
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  53 ++--
 drivers/gpu/drm/i915/i915_gem_gtt.c        |  48 ++--
 drivers/gpu/drm/i915/intel_display.c       |  80 +++---
 drivers/gpu/drm/i915/intel_lrc.c           | 183 +++++++-------
 drivers/gpu/drm/i915/intel_lrc.h           |  26 --
 drivers/gpu/drm/i915/intel_mocs.c          |  38 ++-
 drivers/gpu/drm/i915/intel_overlay.c       |  50 ++--
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 382 +++++++++++++++--------------
 drivers/gpu/drm/i915/intel_ringbuffer.h    |  25 +-
 10 files changed, 450 insertions(+), 482 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index bd13d084e19c..a0e24eb5e167 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -552,6 +552,7 @@ static inline int
 mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 {
 	struct drm_i915_private *dev_priv = req->i915;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_engine_cs *engine = req->engine;
 	u32 flags = hw_flags | MI_MM_SPACE_GTT;
 	const int num_rings =
@@ -589,64 +590,64 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 
 	/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
 	if (INTEL_GEN(dev_priv) >= 7) {
-		intel_ring_emit(engine, MI_ARB_ON_OFF | MI_ARB_DISABLE);
+		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
 		if (num_rings) {
 			struct intel_engine_cs *signaller;
 
-			intel_ring_emit(engine,
+			intel_ring_emit(ring,
 					MI_LOAD_REGISTER_IMM(num_rings));
 			for_each_engine(signaller, dev_priv) {
 				if (signaller == engine)
 					continue;
 
-				intel_ring_emit_reg(engine,
+				intel_ring_emit_reg(ring,
 						    RING_PSMI_CTL(signaller->mmio_base));
-				intel_ring_emit(engine,
+				intel_ring_emit(ring,
 						_MASKED_BIT_ENABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
 			}
 		}
 	}
 
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_emit(engine, MI_SET_CONTEXT);
-	intel_ring_emit(engine,
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_emit(ring, MI_SET_CONTEXT);
+	intel_ring_emit(ring,
 			i915_gem_obj_ggtt_offset(req->ctx->engine[RCS].state) |
 			flags);
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
 	 */
-	intel_ring_emit(engine, MI_NOOP);
+	intel_ring_emit(ring, MI_NOOP);
 
 	if (INTEL_GEN(dev_priv) >= 7) {
 		if (num_rings) {
 			struct intel_engine_cs *signaller;
 			i915_reg_t last_reg = {}; /* keep gcc quiet */
 
-			intel_ring_emit(engine,
+			intel_ring_emit(ring,
 					MI_LOAD_REGISTER_IMM(num_rings));
 			for_each_engine(signaller, dev_priv) {
 				if (signaller == engine)
 					continue;
 
 				last_reg = RING_PSMI_CTL(signaller->mmio_base);
-				intel_ring_emit_reg(engine, last_reg);
-				intel_ring_emit(engine,
+				intel_ring_emit_reg(ring, last_reg);
+				intel_ring_emit(ring,
 						_MASKED_BIT_DISABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
 			}
 
 			/* Insert a delay before the next switch! */
-			intel_ring_emit(engine,
+			intel_ring_emit(ring,
 					MI_STORE_REGISTER_MEM |
 					MI_SRM_LRM_GLOBAL_GTT);
-			intel_ring_emit_reg(engine, last_reg);
-			intel_ring_emit(engine, engine->scratch.gtt_offset);
-			intel_ring_emit(engine, MI_NOOP);
+			intel_ring_emit_reg(ring, last_reg);
+			intel_ring_emit(ring, engine->scratch.gtt_offset);
+			intel_ring_emit(ring, MI_NOOP);
 		}
-		intel_ring_emit(engine, MI_ARB_ON_OFF | MI_ARB_ENABLE);
+		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
 	}
 
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
 	return ret;
 }
@@ -654,7 +655,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 static int remap_l3(struct drm_i915_gem_request *req, int slice)
 {
 	u32 *remap_info = req->i915->l3_parity.remap_info[slice];
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int i, ret;
 
 	if (!remap_info)
@@ -669,13 +670,13 @@ static int remap_l3(struct drm_i915_gem_request *req, int slice)
 	 * here because no other code should access these registers other than
 	 * at initialization time.
 	 */
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(GEN7_L3LOG_SIZE/4));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(GEN7_L3LOG_SIZE/4));
 	for (i = 0; i < GEN7_L3LOG_SIZE/4; i++) {
-		intel_ring_emit_reg(engine, GEN7_L3LOG(slice, i));
-		intel_ring_emit(engine, remap_info[i]);
+		intel_ring_emit_reg(ring, GEN7_L3LOG(slice, i));
+		intel_ring_emit(ring, remap_info[i]);
 	}
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 9ef68e86b483..793128d9410b 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1171,14 +1171,12 @@ i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params)
 }
 
 static int
-i915_reset_gen7_sol_offsets(struct drm_device *dev,
-			    struct drm_i915_gem_request *req)
+i915_reset_gen7_sol_offsets(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret, i;
 
-	if (!IS_GEN7(dev) || engine != &dev_priv->engine[RCS]) {
+	if (!IS_GEN7(req->i915) || req->engine->id != RCS) {
 		DRM_DEBUG("sol reset is gen7/rcs only\n");
 		return -EINVAL;
 	}
@@ -1188,12 +1186,12 @@ i915_reset_gen7_sol_offsets(struct drm_device *dev,
 		return ret;
 
 	for (i = 0; i < 4; i++) {
-		intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(1));
-		intel_ring_emit_reg(engine, GEN7_SO_WRITE_OFFSET(i));
-		intel_ring_emit(engine, 0);
+		intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+		intel_ring_emit_reg(ring, GEN7_SO_WRITE_OFFSET(i));
+		intel_ring_emit(ring, 0);
 	}
 
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1256,9 +1254,7 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 			       struct drm_i915_gem_execbuffer2 *args,
 			       struct list_head *vmas)
 {
-	struct drm_device *dev = params->dev;
-	struct intel_engine_cs *engine = params->engine;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = params->request->i915;
 	u64 exec_start, exec_len;
 	int instp_mode;
 	u32 instp_mask;
@@ -1272,34 +1268,31 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 	if (ret)
 		return ret;
 
-	WARN(params->ctx->ppgtt && params->ctx->ppgtt->pd_dirty_rings & (1<<engine->id),
-	     "%s didn't clear reload\n", engine->name);
-
 	instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK;
 	instp_mask = I915_EXEC_CONSTANTS_MASK;
 	switch (instp_mode) {
 	case I915_EXEC_CONSTANTS_REL_GENERAL:
 	case I915_EXEC_CONSTANTS_ABSOLUTE:
 	case I915_EXEC_CONSTANTS_REL_SURFACE:
-		if (instp_mode != 0 && engine != &dev_priv->engine[RCS]) {
+		if (instp_mode != 0 && params->engine->id != RCS) {
 			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
 			return -EINVAL;
 		}
 
 		if (instp_mode != dev_priv->relative_constants_mode) {
-			if (INTEL_INFO(dev)->gen < 4) {
+			if (INTEL_INFO(dev_priv)->gen < 4) {
 				DRM_DEBUG("no rel constants on pre-gen4\n");
 				return -EINVAL;
 			}
 
-			if (INTEL_INFO(dev)->gen > 5 &&
+			if (INTEL_INFO(dev_priv)->gen > 5 &&
 			    instp_mode == I915_EXEC_CONSTANTS_REL_SURFACE) {
 				DRM_DEBUG("rel surface constants mode invalid on gen5+\n");
 				return -EINVAL;
 			}
 
 			/* The HW changed the meaning on this bit on gen6 */
-			if (INTEL_INFO(dev)->gen >= 6)
+			if (INTEL_INFO(dev_priv)->gen >= 6)
 				instp_mask &= ~I915_EXEC_CONSTANTS_REL_SURFACE;
 		}
 		break;
@@ -1308,23 +1301,25 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 		return -EINVAL;
 	}
 
-	if (engine == &dev_priv->engine[RCS] &&
+	if (params->engine->id == RCS &&
 	    instp_mode != dev_priv->relative_constants_mode) {
+		struct intel_ringbuffer *ring = params->request->ringbuf;
+
 		ret = intel_ring_begin(params->request, 4);
 		if (ret)
 			return ret;
 
-		intel_ring_emit(engine, MI_NOOP);
-		intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(1));
-		intel_ring_emit_reg(engine, INSTPM);
-		intel_ring_emit(engine, instp_mask << 16 | instp_mode);
-		intel_ring_advance(engine);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+		intel_ring_emit_reg(ring, INSTPM);
+		intel_ring_emit(ring, instp_mask << 16 | instp_mode);
+		intel_ring_advance(ring);
 
 		dev_priv->relative_constants_mode = instp_mode;
 	}
 
 	if (args->flags & I915_EXEC_GEN7_SOL_RESET) {
-		ret = i915_reset_gen7_sol_offsets(dev, params->request);
+		ret = i915_reset_gen7_sol_offsets(params->request);
 		if (ret)
 			return ret;
 	}
@@ -1336,9 +1331,9 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 	if (exec_len == 0)
 		exec_len = params->batch_obj->base.size;
 
-	ret = engine->dispatch_execbuffer(params->request,
-					exec_start, exec_len,
-					params->dispatch_flags);
+	ret = params->engine->dispatch_execbuffer(params->request,
+						  exec_start, exec_len,
+						  params->dispatch_flags);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 38e7d992a20d..b38a5311f996 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -669,6 +669,7 @@ static int gen8_write_pdp(struct drm_i915_gem_request *req,
 			  unsigned entry,
 			  dma_addr_t addr)
 {
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -678,13 +679,13 @@ static int gen8_write_pdp(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(1));
-	intel_ring_emit_reg(engine, GEN8_RING_PDP_UDW(engine, entry));
-	intel_ring_emit(engine, upper_32_bits(addr));
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(1));
-	intel_ring_emit_reg(engine, GEN8_RING_PDP_LDW(engine, entry));
-	intel_ring_emit(engine, lower_32_bits(addr));
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+	intel_ring_emit_reg(ring, GEN8_RING_PDP_UDW(engine, entry));
+	intel_ring_emit(ring, upper_32_bits(addr));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+	intel_ring_emit_reg(ring, GEN8_RING_PDP_LDW(engine, entry));
+	intel_ring_emit(ring, lower_32_bits(addr));
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1660,6 +1661,7 @@ static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			 struct drm_i915_gem_request *req)
 {
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1672,13 +1674,13 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(2));
-	intel_ring_emit_reg(engine, RING_PP_DIR_DCLV(engine));
-	intel_ring_emit(engine, PP_DIR_DCLV_2G);
-	intel_ring_emit_reg(engine, RING_PP_DIR_BASE(engine));
-	intel_ring_emit(engine, get_pd_offset(ppgtt));
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(2));
+	intel_ring_emit_reg(ring, RING_PP_DIR_DCLV(engine));
+	intel_ring_emit(ring, PP_DIR_DCLV_2G);
+	intel_ring_emit_reg(ring, RING_PP_DIR_BASE(engine));
+	intel_ring_emit(ring, get_pd_offset(ppgtt));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1686,6 +1688,7 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			  struct drm_i915_gem_request *req)
 {
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1698,17 +1701,18 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(2));
-	intel_ring_emit_reg(engine, RING_PP_DIR_DCLV(engine));
-	intel_ring_emit(engine, PP_DIR_DCLV_2G);
-	intel_ring_emit_reg(engine, RING_PP_DIR_BASE(engine));
-	intel_ring_emit(engine, get_pd_offset(ppgtt));
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(2));
+	intel_ring_emit_reg(ring, RING_PP_DIR_DCLV(engine));
+	intel_ring_emit(ring, PP_DIR_DCLV_2G);
+	intel_ring_emit_reg(ring, RING_PP_DIR_BASE(engine));
+	intel_ring_emit(ring, get_pd_offset(ppgtt));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	/* XXX: RCS is the only one to auto invalidate the TLBs? */
 	if (engine->id != RCS) {
-		ret = engine->flush(req, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+		ret = engine->flush(req,
+				    I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 78beb7e9d384..d18ed32e6a31 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11123,7 +11123,7 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11139,13 +11139,13 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 		flip_mask = MI_WAIT_FOR_PLANE_B_FLIP;
 	else
 		flip_mask = MI_WAIT_FOR_PLANE_A_FLIP;
-	intel_ring_emit(engine, MI_WAIT_FOR_EVENT | flip_mask);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_emit(engine, MI_DISPLAY_FLIP |
+	intel_ring_emit(ring, MI_WAIT_FOR_EVENT | flip_mask);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
-	intel_ring_emit(engine, fb->pitches[0]);
-	intel_ring_emit(engine, intel_crtc->flip_work->gtt_offset);
-	intel_ring_emit(engine, 0); /* aux display base address, unused */
+	intel_ring_emit(ring, fb->pitches[0]);
+	intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
+	intel_ring_emit(ring, 0); /* aux display base address, unused */
 
 	return 0;
 }
@@ -11157,7 +11157,7 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11170,13 +11170,13 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 		flip_mask = MI_WAIT_FOR_PLANE_B_FLIP;
 	else
 		flip_mask = MI_WAIT_FOR_PLANE_A_FLIP;
-	intel_ring_emit(engine, MI_WAIT_FOR_EVENT | flip_mask);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_emit(engine, MI_DISPLAY_FLIP_I915 |
+	intel_ring_emit(ring, MI_WAIT_FOR_EVENT | flip_mask);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
-	intel_ring_emit(engine, fb->pitches[0]);
-	intel_ring_emit(engine, intel_crtc->flip_work->gtt_offset);
-	intel_ring_emit(engine, MI_NOOP);
+	intel_ring_emit(ring, fb->pitches[0]);
+	intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
+	intel_ring_emit(ring, MI_NOOP);
 
 	return 0;
 }
@@ -11188,7 +11188,7 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11202,10 +11202,10 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 	 * Display Registers (which do not change across a page-flip)
 	 * so we need only reprogram the base address.
 	 */
-	intel_ring_emit(engine, MI_DISPLAY_FLIP |
+	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
-	intel_ring_emit(engine, fb->pitches[0]);
-	intel_ring_emit(engine, intel_crtc->flip_work->gtt_offset |
+	intel_ring_emit(ring, fb->pitches[0]);
+	intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset |
 			obj->tiling_mode);
 
 	/* XXX Enabling the panel-fitter across page-flip is so far
@@ -11214,7 +11214,7 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 	 */
 	pf = 0;
 	pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
-	intel_ring_emit(engine, pf | pipesrc);
+	intel_ring_emit(ring, pf | pipesrc);
 
 	return 0;
 }
@@ -11226,7 +11226,7 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11236,10 +11236,10 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_DISPLAY_FLIP |
+	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
-	intel_ring_emit(engine, fb->pitches[0] | obj->tiling_mode);
-	intel_ring_emit(engine, intel_crtc->flip_work->gtt_offset);
+	intel_ring_emit(ring, fb->pitches[0] | obj->tiling_mode);
+	intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
 
 	/* Contrary to the suggestions in the documentation,
 	 * "Enable Panel Fitter" does not seem to be required when page
@@ -11249,7 +11249,7 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 	 */
 	pf = 0;
 	pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
-	intel_ring_emit(engine, pf | pipesrc);
+	intel_ring_emit(ring, pf | pipesrc);
 
 	return 0;
 }
@@ -11261,7 +11261,7 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t plane_bit = 0;
 	int len, ret;
@@ -11282,7 +11282,7 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 	}
 
 	len = 4;
-	if (engine->id == RCS) {
+	if (req->engine->id == RCS) {
 		len += 6;
 		/*
 		 * On Gen 8, SRM is now taking an extra dword to accommodate
@@ -11320,30 +11320,30 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 	 * for the RCS also doesn't appear to drop events. Setting the DERRMR
 	 * to zero does lead to lockups within MI_DISPLAY_FLIP.
 	 */
-	if (engine->id == RCS) {
-		intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(1));
-		intel_ring_emit_reg(engine, DERRMR);
-		intel_ring_emit(engine, ~(DERRMR_PIPEA_PRI_FLIP_DONE |
+	if (req->engine->id == RCS) {
+		intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+		intel_ring_emit_reg(ring, DERRMR);
+		intel_ring_emit(ring, ~(DERRMR_PIPEA_PRI_FLIP_DONE |
 					  DERRMR_PIPEB_PRI_FLIP_DONE |
 					  DERRMR_PIPEC_PRI_FLIP_DONE));
 		if (IS_GEN8(dev))
-			intel_ring_emit(engine, MI_STORE_REGISTER_MEM_GEN8 |
+			intel_ring_emit(ring, MI_STORE_REGISTER_MEM_GEN8 |
 					      MI_SRM_LRM_GLOBAL_GTT);
 		else
-			intel_ring_emit(engine, MI_STORE_REGISTER_MEM |
+			intel_ring_emit(ring, MI_STORE_REGISTER_MEM |
 					      MI_SRM_LRM_GLOBAL_GTT);
-		intel_ring_emit_reg(engine, DERRMR);
-		intel_ring_emit(engine, engine->scratch.gtt_offset + 256);
+		intel_ring_emit_reg(ring, DERRMR);
+		intel_ring_emit(ring, req->engine->scratch.gtt_offset + 256);
 		if (IS_GEN8(dev)) {
-			intel_ring_emit(engine, 0);
-			intel_ring_emit(engine, MI_NOOP);
+			intel_ring_emit(ring, 0);
+			intel_ring_emit(ring, MI_NOOP);
 		}
 	}
 
-	intel_ring_emit(engine, MI_DISPLAY_FLIP_I915 | plane_bit);
-	intel_ring_emit(engine, (fb->pitches[0] | obj->tiling_mode));
-	intel_ring_emit(engine, intel_crtc->flip_work->gtt_offset);
-	intel_ring_emit(engine, (MI_NOOP));
+	intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 | plane_bit);
+	intel_ring_emit(ring, (fb->pitches[0] | obj->tiling_mode));
+	intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
+	intel_ring_emit(ring, (MI_NOOP));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index daf1279a317d..1db2fdadbb46 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -773,7 +773,7 @@ intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 	struct intel_ringbuffer *ringbuf = request->ringbuf;
 	struct intel_engine_cs *engine = request->engine;
 
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_advance(ringbuf);
 	request->tail = ringbuf->tail;
 
 	/*
@@ -782,9 +782,9 @@ intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 	 *
 	 * Caller must reserve WA_TAIL_DWORDS for us!
 	 */
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_advance(ringbuf);
 
 	/* We keep the previous context alive until we retire the following
 	 * request. This ensures that any the context object is still pinned
@@ -868,11 +868,11 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 		if (ret)
 			return ret;
 
-		intel_logical_ring_emit(ringbuf, MI_NOOP);
-		intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
-		intel_logical_ring_emit_reg(ringbuf, INSTPM);
-		intel_logical_ring_emit(ringbuf, instp_mask << 16 | instp_mode);
-		intel_logical_ring_advance(ringbuf);
+		intel_ring_emit(ringbuf, MI_NOOP);
+		intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
+		intel_ring_emit_reg(ringbuf, INSTPM);
+		intel_ring_emit(ringbuf, instp_mask << 16 | instp_mode);
+		intel_ring_advance(ringbuf);
 
 		dev_priv->relative_constants_mode = instp_mode;
 	}
@@ -1045,14 +1045,14 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(w->count));
+	intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(w->count));
 	for (i = 0; i < w->count; i++) {
-		intel_logical_ring_emit_reg(ringbuf, w->reg[i].addr);
-		intel_logical_ring_emit(ringbuf, w->reg[i].value);
+		intel_ring_emit_reg(ringbuf, w->reg[i].addr);
+		intel_ring_emit(ringbuf, w->reg[i].value);
 	}
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_emit(ringbuf, MI_NOOP);
 
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_advance(ringbuf);
 
 	engine->gpu_caches_dirty = true;
 	ret = logical_ring_flush_all_caches(req);
@@ -1553,8 +1553,8 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine)
 static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 {
 	struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
 	const int num_lri_cmds = GEN8_LEGACY_PDPES * 2;
 	int i, ret;
 
@@ -1562,20 +1562,18 @@ static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(num_lri_cmds));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_lri_cmds));
 	for (i = GEN8_LEGACY_PDPES - 1; i >= 0; i--) {
 		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
 
-		intel_logical_ring_emit_reg(ringbuf,
-					    GEN8_RING_PDP_UDW(engine, i));
-		intel_logical_ring_emit(ringbuf, upper_32_bits(pd_daddr));
-		intel_logical_ring_emit_reg(ringbuf,
-					    GEN8_RING_PDP_LDW(engine, i));
-		intel_logical_ring_emit(ringbuf, lower_32_bits(pd_daddr));
+		intel_ring_emit_reg(ring, GEN8_RING_PDP_UDW(engine, i));
+		intel_ring_emit(ring, upper_32_bits(pd_daddr));
+		intel_ring_emit_reg(ring, GEN8_RING_PDP_LDW(engine, i));
+		intel_ring_emit(ring, lower_32_bits(pd_daddr));
 	}
 
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1583,7 +1581,7 @@ static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
 			      u64 offset, unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	bool ppgtt = !(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
 
@@ -1610,14 +1608,14 @@ static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
 		return ret;
 
 	/* FIXME(BDW): Address space and security selectors. */
-	intel_logical_ring_emit(ringbuf, MI_BATCH_BUFFER_START_GEN8 |
-				(ppgtt<<8) |
-				(dispatch_flags & I915_DISPATCH_RS ?
-				 MI_BATCH_RESOURCE_STREAMER : 0));
-	intel_logical_ring_emit(ringbuf, lower_32_bits(offset));
-	intel_logical_ring_emit(ringbuf, upper_32_bits(offset));
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ring, MI_BATCH_BUFFER_START_GEN8 |
+			(ppgtt<<8) |
+			(dispatch_flags & I915_DISPATCH_RS ?
+			 MI_BATCH_RESOURCE_STREAMER : 0));
+	intel_ring_emit(ring, lower_32_bits(offset));
+	intel_ring_emit(ring, upper_32_bits(offset));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1640,9 +1638,8 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 			   u32 invalidate_domains,
 			   u32 unused)
 {
-	struct intel_ringbuffer *ringbuf = request->ringbuf;
-	struct intel_engine_cs *engine = ringbuf->engine;
-	struct drm_i915_private *dev_priv = request->i915;
+	struct intel_ringbuffer *ring = request->ringbuf;
+	struct intel_engine_cs *engine = ring->engine;
 	uint32_t cmd;
 	int ret;
 
@@ -1661,17 +1658,17 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 
 	if (invalidate_domains & I915_GEM_GPU_DOMAINS) {
 		cmd |= MI_INVALIDATE_TLB;
-		if (engine == &dev_priv->engine[VCS])
+		if (engine->id == VCS)
 			cmd |= MI_INVALIDATE_BSD;
 	}
 
-	intel_logical_ring_emit(ringbuf, cmd);
-	intel_logical_ring_emit(ringbuf,
-				I915_GEM_HWS_SCRATCH_ADDR |
-				MI_FLUSH_DW_USE_GTT);
-	intel_logical_ring_emit(ringbuf, 0); /* upper addr */
-	intel_logical_ring_emit(ringbuf, 0); /* value */
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ring, cmd);
+	intel_ring_emit(ring,
+			I915_GEM_HWS_SCRATCH_ADDR |
+			MI_FLUSH_DW_USE_GTT);
+	intel_ring_emit(ring, 0); /* upper addr */
+	intel_ring_emit(ring, 0); /* value */
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1680,8 +1677,8 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 				  u32 invalidate_domains,
 				  u32 flush_domains)
 {
-	struct intel_ringbuffer *ringbuf = request->ringbuf;
-	struct intel_engine_cs *engine = ringbuf->engine;
+	struct intel_ringbuffer *ring = request->ringbuf;
+	struct intel_engine_cs *engine = request->engine;
 	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	bool vf_flush_wa = false, dc_flush_wa = false;
 	u32 flags = 0;
@@ -1732,40 +1729,40 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 		return ret;
 
 	if (vf_flush_wa) {
-		intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
+		intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
 	}
 
 	if (dc_flush_wa) {
-		intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
-		intel_logical_ring_emit(ringbuf, PIPE_CONTROL_DC_FLUSH_ENABLE);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
+		intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+		intel_ring_emit(ring, PIPE_CONTROL_DC_FLUSH_ENABLE);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
 	}
 
-	intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
-	intel_logical_ring_emit(ringbuf, flags);
-	intel_logical_ring_emit(ringbuf, scratch_addr);
-	intel_logical_ring_emit(ringbuf, 0);
-	intel_logical_ring_emit(ringbuf, 0);
-	intel_logical_ring_emit(ringbuf, 0);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+	intel_ring_emit(ring, flags);
+	intel_ring_emit(ring, scratch_addr);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
 
 	if (dc_flush_wa) {
-		intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
-		intel_logical_ring_emit(ringbuf, PIPE_CONTROL_CS_STALL);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
-		intel_logical_ring_emit(ringbuf, 0);
+		intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+		intel_ring_emit(ring, PIPE_CONTROL_CS_STALL);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, 0);
 	}
 
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1794,7 +1791,7 @@ static void bxt_a_seqno_barrier(struct intel_engine_cs *engine)
 
 static int gen8_emit_request(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ringbuf = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(request, 6 + WA_TAIL_DWORDS);
@@ -1804,21 +1801,20 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)
 	/* w/a: bit 5 needs to be zero for MI_FLUSH_DW address. */
 	BUILD_BUG_ON(I915_GEM_HWS_INDEX_ADDR & (1 << 5));
 
-	intel_logical_ring_emit(ringbuf,
-				(MI_FLUSH_DW + 1) | MI_FLUSH_DW_OP_STOREDW);
-	intel_logical_ring_emit(ringbuf,
-				intel_hws_seqno_address(request->engine) |
-				MI_FLUSH_DW_USE_GTT);
-	intel_logical_ring_emit(ringbuf, 0);
-	intel_logical_ring_emit(ringbuf, request->fence.seqno);
-	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_emit(ring, (MI_FLUSH_DW + 1) | MI_FLUSH_DW_OP_STOREDW);
+	intel_ring_emit(ring,
+			intel_hws_seqno_address(request->engine) |
+			MI_FLUSH_DW_USE_GTT);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, request->fence.seqno);
+	intel_ring_emit(ring, MI_USER_INTERRUPT);
+	intel_ring_emit(ring, MI_NOOP);
 	return intel_logical_ring_advance_and_submit(request);
 }
 
 static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ringbuf = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(request, 8 + WA_TAIL_DWORDS);
@@ -1832,19 +1828,18 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 	 * need a prior CS_STALL, which is emitted by the flush
 	 * following the batch.
 	 */
-	intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
-	intel_logical_ring_emit(ringbuf,
-				(PIPE_CONTROL_GLOBAL_GTT_IVB |
-				 PIPE_CONTROL_CS_STALL |
-				 PIPE_CONTROL_QW_WRITE));
-	intel_logical_ring_emit(ringbuf,
-				intel_hws_seqno_address(request->engine));
-	intel_logical_ring_emit(ringbuf, 0);
-	intel_logical_ring_emit(ringbuf, i915_gem_request_get_seqno(request));
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+	intel_ring_emit(ring,
+			(PIPE_CONTROL_GLOBAL_GTT_IVB |
+			 PIPE_CONTROL_CS_STALL |
+			 PIPE_CONTROL_QW_WRITE));
+	intel_ring_emit(ring, intel_hws_seqno_address(request->engine));
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, i915_gem_request_get_seqno(request));
 	/* We're thrashing one dword of HWS. */
-	intel_logical_ring_emit(ringbuf, 0);
-	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, MI_USER_INTERRUPT);
+	intel_ring_emit(ring, MI_NOOP);
 	return intel_logical_ring_advance_and_submit(request);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index 38287302c5ba..d26fb44549e5 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -73,32 +73,6 @@ int logical_xcs_ring_init(struct intel_engine_cs *engine);
 int intel_engines_init(struct drm_device *dev);
 
 int logical_ring_flush_all_caches(struct drm_i915_gem_request *req);
-/**
- * intel_logical_ring_advance() - advance the ringbuffer tail
- * @ringbuf: Ringbuffer to advance.
- *
- * The tail is only updated in our logical ringbuffer struct.
- */
-static inline void intel_logical_ring_advance(struct intel_ringbuffer *ringbuf)
-{
-	__intel_ringbuffer_advance(ringbuf);
-}
-
-/**
- * intel_logical_ring_emit() - write a DWORD to the ringbuffer.
- * @ringbuf: Ringbuffer to write to.
- * @data: DWORD to write.
- */
-static inline void intel_logical_ring_emit(struct intel_ringbuffer *ringbuf,
-					   u32 data)
-{
-	__intel_ringbuffer_emit(ringbuf, data);
-}
-static inline void intel_logical_ring_emit_reg(struct intel_ringbuffer *ringbuf,
-					       i915_reg_t reg)
-{
-	intel_logical_ring_emit(ringbuf, i915_mmio_reg_offset(reg));
-}
 
 /* Logical Ring Contexts */
 
diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c
index bd46968d8a07..3059c52030b4 100644
--- a/drivers/gpu/drm/i915/intel_mocs.c
+++ b/drivers/gpu/drm/i915/intel_mocs.c
@@ -288,14 +288,11 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_logical_ring_emit(ringbuf,
-				MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES));
+	intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES));
 
 	for (index = 0; index < table->size; index++) {
-		intel_logical_ring_emit_reg(ringbuf,
-					    mocs_register(engine, index));
-		intel_logical_ring_emit(ringbuf,
-					table->table[index].control_value);
+		intel_ring_emit_reg(ringbuf, mocs_register(engine, index));
+		intel_ring_emit(ringbuf, table->table[index].control_value);
 	}
 
 	/*
@@ -307,14 +304,12 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 	 * that value to all the used entries.
 	 */
 	for (; index < GEN9_NUM_MOCS_ENTRIES; index++) {
-		intel_logical_ring_emit_reg(ringbuf,
-					    mocs_register(engine, index));
-		intel_logical_ring_emit(ringbuf,
-					table->table[0].control_value);
+		intel_ring_emit_reg(ringbuf, mocs_register(engine, index));
+		intel_ring_emit(ringbuf, table->table[0].control_value);
 	}
 
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_advance(ringbuf);
 
 	return 0;
 }
@@ -352,19 +347,18 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_logical_ring_emit(ringbuf,
+	intel_ring_emit(ringbuf,
 			MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES / 2));
 
 	for (i = 0; i < table->size/2; i++) {
-		intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_logical_ring_emit(ringbuf,
-					l3cc_combine(table, 2*i, 2*i+1));
+		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ringbuf, l3cc_combine(table, 2*i, 2*i+1));
 	}
 
 	if (table->size & 0x01) {
 		/* Odd table size - 1 left over */
-		intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_logical_ring_emit(ringbuf, l3cc_combine(table, 2*i, 0));
+		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ringbuf, l3cc_combine(table, 2*i, 0));
 		i++;
 	}
 
@@ -374,12 +368,12 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 	 * they are reserved by the hardware.
 	 */
 	for (; i < GEN9_NUM_MOCS_ENTRIES / 2; i++) {
-		intel_logical_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_logical_ring_emit(ringbuf, l3cc_combine(table, 0, 0));
+		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ringbuf, l3cc_combine(table, 0, 0));
 	}
 
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
-	intel_logical_ring_advance(ringbuf);
+	intel_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_advance(ringbuf);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index c10ce368787e..ec63b64fb202 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -235,6 +235,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
+	struct intel_ringbuffer *ring;
 	int ret;
 
 	WARN_ON(overlay->active);
@@ -252,11 +253,12 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 
 	overlay->active = true;
 
-	intel_ring_emit(engine, MI_OVERLAY_FLIP | MI_OVERLAY_ON);
-	intel_ring_emit(engine, overlay->flip_addr | OFC_UPDATE);
-	intel_ring_emit(engine, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	ring = req->ringbuf;
+	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_ON);
+	intel_ring_emit(ring, overlay->flip_addr | OFC_UPDATE);
+	intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return intel_overlay_do_wait_request(overlay, req, NULL);
 }
@@ -268,6 +270,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
+	struct intel_ringbuffer *ring;
 	u32 flip_addr = overlay->flip_addr;
 	u32 tmp;
 	int ret;
@@ -292,9 +295,10 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 		return ret;
 	}
 
-	intel_ring_emit(engine, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
-	intel_ring_emit(engine, flip_addr);
-	intel_ring_advance(engine);
+	ring = req->ringbuf;
+	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
+	intel_ring_emit(ring, flip_addr);
+	intel_ring_advance(ring);
 
 	WARN_ON(overlay->last_flip_req);
 	i915_gem_request_assign(&overlay->last_flip_req, req);
@@ -336,6 +340,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
+	struct intel_ringbuffer *ring;
 	u32 flip_addr = overlay->flip_addr;
 	int ret;
 
@@ -357,24 +362,25 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 		return ret;
 	}
 
+	ring = req->ringbuf;
 	/* wait for overlay to go idle */
-	intel_ring_emit(engine, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
-	intel_ring_emit(engine, flip_addr);
-	intel_ring_emit(engine, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
+	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
+	intel_ring_emit(ring, flip_addr);
+	intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
 	/* turn overlay off */
 	if (IS_I830(dev_priv)) {
 		/* Workaround: Don't disable the overlay fully, since otherwise
 		 * it dies on the next OVERLAY_ON cmd. */
-		intel_ring_emit(engine, MI_NOOP);
-		intel_ring_emit(engine, MI_NOOP);
-		intel_ring_emit(engine, MI_NOOP);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_emit(ring, MI_NOOP);
 	} else {
-		intel_ring_emit(engine, MI_OVERLAY_FLIP | MI_OVERLAY_OFF);
-		intel_ring_emit(engine, flip_addr);
-		intel_ring_emit(engine,
+		intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_OFF);
+		intel_ring_emit(ring, flip_addr);
+		intel_ring_emit(ring,
 				MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
 	}
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
 	return intel_overlay_do_wait_request(overlay, req, intel_overlay_off_tail);
 }
@@ -420,6 +426,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 	if (I915_READ(ISR) & I915_OVERLAY_PLANE_FLIP_PENDING_INTERRUPT) {
 		/* synchronous slowpath */
 		struct drm_i915_gem_request *req;
+		struct intel_ringbuffer *ring;
 
 		req = i915_gem_request_alloc(engine, NULL);
 		if (IS_ERR(req))
@@ -431,10 +438,11 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 			return ret;
 		}
 
-		intel_ring_emit(engine,
+		ring = req->ringbuf;
+		intel_ring_emit(ring,
 				MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
-		intel_ring_emit(engine, MI_NOOP);
-		intel_ring_advance(engine);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_advance(ring);
 
 		ret = intel_overlay_do_wait_request(overlay, req,
 						    intel_overlay_release_old_vid_tail);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 0b5d1de8a7fb..48cdfbec1f46 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -58,7 +58,7 @@ void intel_ring_update_space(struct intel_ringbuffer *ringbuf)
 					    ringbuf->tail, ringbuf->size);
 }
 
-static void __intel_ring_advance(struct intel_engine_cs *engine)
+static void __intel_engine_submit(struct intel_engine_cs *engine)
 {
 	struct intel_ringbuffer *ringbuf = engine->buffer;
 	ringbuf->tail &= ringbuf->size - 1;
@@ -70,7 +70,7 @@ gen2_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	u32 cmd;
 	int ret;
 
@@ -85,9 +85,9 @@ gen2_render_ring_flush(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, cmd);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, cmd);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -97,7 +97,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	u32 cmd;
 	int ret;
 
@@ -129,23 +129,20 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 	 * are flushed at any MI_FLUSH.
 	 */
 
-	cmd = MI_FLUSH | MI_NO_WRITE_FLUSH;
-	if ((invalidate_domains|flush_domains) & I915_GEM_DOMAIN_RENDER)
-		cmd &= ~MI_NO_WRITE_FLUSH;
-	if (invalidate_domains & I915_GEM_DOMAIN_INSTRUCTION)
+	cmd = MI_FLUSH;
+	if (invalidate_domains) {
 		cmd |= MI_EXE_FLUSH;
-
-	if (invalidate_domains & I915_GEM_DOMAIN_COMMAND &&
-	    (IS_G4X(req->i915) || IS_GEN5(req->i915)))
-		cmd |= MI_INVALIDATE_ISP;
+		if (IS_G4X(req->i915) || IS_GEN5(req->i915))
+			cmd |= MI_INVALIDATE_ISP;
+	}
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, cmd);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, cmd);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -190,34 +187,35 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 intel_emit_post_sync_nonzero_flush(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
-	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+	struct intel_ringbuffer *ring = req->ringbuf;
+	u32 scratch_addr =
+		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	ret = intel_ring_begin(req, 6);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(5));
-	intel_ring_emit(engine, PIPE_CONTROL_CS_STALL |
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(5));
+	intel_ring_emit(ring, PIPE_CONTROL_CS_STALL |
 			PIPE_CONTROL_STALL_AT_SCOREBOARD);
-	intel_ring_emit(engine, scratch_addr | PIPE_CONTROL_GLOBAL_GTT); /* address */
-	intel_ring_emit(engine, 0); /* low dword */
-	intel_ring_emit(engine, 0); /* high dword */
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT);
+	intel_ring_emit(ring, 0); /* low dword */
+	intel_ring_emit(ring, 0); /* high dword */
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	ret = intel_ring_begin(req, 6);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(5));
-	intel_ring_emit(engine, PIPE_CONTROL_QW_WRITE);
-	intel_ring_emit(engine, scratch_addr | PIPE_CONTROL_GLOBAL_GTT); /* address */
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(5));
+	intel_ring_emit(ring, PIPE_CONTROL_QW_WRITE);
+	intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -226,9 +224,10 @@ static int
 gen6_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
+	u32 scratch_addr =
+		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
-	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	/* Force SNB workarounds for PIPE_CONTROL flushes */
@@ -266,11 +265,11 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(4));
-	intel_ring_emit(engine, flags);
-	intel_ring_emit(engine, scratch_addr | PIPE_CONTROL_GLOBAL_GTT);
-	intel_ring_emit(engine, 0);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4));
+	intel_ring_emit(ring, flags);
+	intel_ring_emit(ring, scratch_addr | PIPE_CONTROL_GLOBAL_GTT);
+	intel_ring_emit(ring, 0);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -278,19 +277,20 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 gen7_render_ring_cs_stall_wa(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(4));
-	intel_ring_emit(engine, PIPE_CONTROL_CS_STALL |
-			      PIPE_CONTROL_STALL_AT_SCOREBOARD);
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, 0);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4));
+	intel_ring_emit(ring,
+			PIPE_CONTROL_CS_STALL |
+			PIPE_CONTROL_STALL_AT_SCOREBOARD);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -299,9 +299,10 @@ static int
 gen7_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
+	u32 scratch_addr =
+		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
-	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	/*
@@ -350,11 +351,11 @@ gen7_render_ring_flush(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(4));
-	intel_ring_emit(engine, flags);
-	intel_ring_emit(engine, scratch_addr);
-	intel_ring_emit(engine, 0);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4));
+	intel_ring_emit(ring, flags);
+	intel_ring_emit(ring, scratch_addr);
+	intel_ring_emit(ring, 0);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -363,20 +364,20 @@ static int
 gen8_emit_pipe_control(struct drm_i915_gem_request *req,
 		       u32 flags, u32 scratch_addr)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 6);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(6));
-	intel_ring_emit(engine, flags);
-	intel_ring_emit(engine, scratch_addr);
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, 0);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+	intel_ring_emit(ring, flags);
+	intel_ring_emit(ring, scratch_addr);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, 0);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -385,8 +386,8 @@ static int
 gen8_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	u32 flags = 0;
 	u32 scratch_addr = req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+	u32 flags = 0;
 	int ret;
 
 	flags |= PIPE_CONTROL_CS_STALL;
@@ -679,14 +680,14 @@ err:
 
 static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	struct i915_workarounds *w = &req->i915->workarounds;
 	int ret, i;
 
 	if (w->count == 0)
 		return 0;
 
-	engine->gpu_caches_dirty = true;
+	req->engine->gpu_caches_dirty = true;
 	ret = intel_ring_flush_all_caches(req);
 	if (ret)
 		return ret;
@@ -695,16 +696,16 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_LOAD_REGISTER_IMM(w->count));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(w->count));
 	for (i = 0; i < w->count; i++) {
-		intel_ring_emit_reg(engine, w->reg[i].addr);
-		intel_ring_emit(engine, w->reg[i].value);
+		intel_ring_emit_reg(ring, w->reg[i].addr);
+		intel_ring_emit(ring, w->reg[i].value);
 	}
-	intel_ring_emit(engine, MI_NOOP);
+	intel_ring_emit(ring, MI_NOOP);
 
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
-	engine->gpu_caches_dirty = true;
+	req->engine->gpu_caches_dirty = true;
 	ret = intel_ring_flush_all_caches(req);
 	if (ret)
 		return ret;
@@ -1337,7 +1338,7 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 8
-	struct intel_engine_cs *signaller = signaller_req->engine;
+	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1352,20 +1353,23 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 		return ret;
 
 	for_each_engine_id(waiter, dev_priv, id) {
-		u64 gtt_offset = signaller->semaphore.signal_ggtt[id];
+		u64 gtt_offset =
+			signaller_req->engine->semaphore.signal_ggtt[id];
 		if (gtt_offset == MI_SEMAPHORE_SYNC_INVALID)
 			continue;
 
 		intel_ring_emit(signaller, GFX_OP_PIPE_CONTROL(6));
-		intel_ring_emit(signaller, PIPE_CONTROL_GLOBAL_GTT_IVB |
-					   PIPE_CONTROL_QW_WRITE |
-					   PIPE_CONTROL_CS_STALL);
+		intel_ring_emit(signaller,
+				PIPE_CONTROL_GLOBAL_GTT_IVB |
+				PIPE_CONTROL_QW_WRITE |
+				PIPE_CONTROL_CS_STALL);
 		intel_ring_emit(signaller, lower_32_bits(gtt_offset));
 		intel_ring_emit(signaller, upper_32_bits(gtt_offset));
 		intel_ring_emit(signaller, signaller_req->fence.seqno);
 		intel_ring_emit(signaller, 0);
-		intel_ring_emit(signaller, MI_SEMAPHORE_SIGNAL |
-					   MI_SEMAPHORE_TARGET(waiter->hw_id));
+		intel_ring_emit(signaller,
+				MI_SEMAPHORE_SIGNAL |
+				MI_SEMAPHORE_TARGET(waiter->hw_id));
 		intel_ring_emit(signaller, 0);
 	}
 
@@ -1376,7 +1380,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 6
-	struct intel_engine_cs *signaller = signaller_req->engine;
+	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1391,18 +1395,21 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 		return ret;
 
 	for_each_engine_id(waiter, dev_priv, id) {
-		u64 gtt_offset = signaller->semaphore.signal_ggtt[id];
+		u64 gtt_offset =
+			signaller_req->engine->semaphore.signal_ggtt[id];
 		if (gtt_offset == MI_SEMAPHORE_SYNC_INVALID)
 			continue;
 
-		intel_ring_emit(signaller, (MI_FLUSH_DW + 1) |
-					   MI_FLUSH_DW_OP_STOREDW);
-		intel_ring_emit(signaller, lower_32_bits(gtt_offset) |
-					   MI_FLUSH_DW_USE_GTT);
+		intel_ring_emit(signaller,
+				(MI_FLUSH_DW + 1) | MI_FLUSH_DW_OP_STOREDW);
+		intel_ring_emit(signaller,
+				lower_32_bits(gtt_offset) |
+				MI_FLUSH_DW_USE_GTT);
 		intel_ring_emit(signaller, upper_32_bits(gtt_offset));
 		intel_ring_emit(signaller, signaller_req->fence.seqno);
-		intel_ring_emit(signaller, MI_SEMAPHORE_SIGNAL |
-					   MI_SEMAPHORE_TARGET(waiter->hw_id));
+		intel_ring_emit(signaller,
+				MI_SEMAPHORE_SIGNAL |
+				MI_SEMAPHORE_TARGET(waiter->hw_id));
 		intel_ring_emit(signaller, 0);
 	}
 
@@ -1412,7 +1419,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 		       unsigned int num_dwords)
 {
-	struct intel_engine_cs *signaller = signaller_req->engine;
+	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *useless;
 	enum intel_engine_id id;
@@ -1428,7 +1435,8 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 		return ret;
 
 	for_each_engine_id(useless, dev_priv, id) {
-		i915_reg_t mbox_reg = signaller->semaphore.mbox.signal[id];
+		i915_reg_t mbox_reg =
+			signaller_req->engine->semaphore.mbox.signal[id];
 
 		if (i915_mmio_reg_valid(mbox_reg)) {
 			intel_ring_emit(signaller, MI_LOAD_REGISTER_IMM(1));
@@ -1456,6 +1464,7 @@ static int
 gen6_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1466,12 +1475,11 @@ gen6_add_request(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_STORE_DWORD_INDEX);
-	intel_ring_emit(engine,
-			I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-	intel_ring_emit(engine, req->fence.seqno);
-	intel_ring_emit(engine, MI_USER_INTERRUPT);
-	__intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
+	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
+	intel_ring_emit(ring, req->fence.seqno);
+	intel_ring_emit(ring, MI_USER_INTERRUPT);
+	__intel_engine_submit(engine);
 
 	return 0;
 }
@@ -1480,6 +1488,7 @@ static int
 gen8_render_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1489,18 +1498,18 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, GFX_OP_PIPE_CONTROL(6));
-	intel_ring_emit(engine, (PIPE_CONTROL_GLOBAL_GTT_IVB |
-				 PIPE_CONTROL_CS_STALL |
-				 PIPE_CONTROL_QW_WRITE));
-	intel_ring_emit(engine, intel_hws_seqno_address(req->engine));
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, i915_gem_request_get_seqno(req));
+	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+	intel_ring_emit(ring, (PIPE_CONTROL_GLOBAL_GTT_IVB |
+			       PIPE_CONTROL_CS_STALL |
+			       PIPE_CONTROL_QW_WRITE));
+	intel_ring_emit(ring, intel_hws_seqno_address(engine));
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, i915_gem_request_get_seqno(req));
 	/* We're thrashing one dword of HWS. */
-	intel_ring_emit(engine, 0);
-	intel_ring_emit(engine, MI_USER_INTERRUPT);
-	intel_ring_emit(engine, MI_NOOP);
-	__intel_ring_advance(engine);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, MI_USER_INTERRUPT);
+	intel_ring_emit(ring, MI_NOOP);
+	__intel_engine_submit(engine);
 
 	return 0;
 }
@@ -1524,9 +1533,9 @@ gen8_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_engine_cs *waiter = waiter_req->engine;
+	struct intel_ringbuffer *waiter = waiter_req->ringbuf;
 	struct drm_i915_private *dev_priv = waiter_req->i915;
-	u64 offset = GEN8_WAIT_OFFSET(waiter, signaller->id);
+	u64 offset = GEN8_WAIT_OFFSET(waiter_req->engine, signaller->id);
 	struct i915_hw_ppgtt *ppgtt;
 	int ret;
 
@@ -1558,11 +1567,11 @@ gen6_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_engine_cs *waiter = waiter_req->engine;
+	struct intel_ringbuffer *waiter = waiter_req->ringbuf;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
-	u32 wait_mbox = signaller->semaphore.mbox.wait[waiter->id];
+	u32 wait_mbox = signaller->semaphore.mbox.wait[waiter_req->engine->id];
 	int ret;
 
 	/* Throughout all of the GEM code, seqno passed implies our current
@@ -1692,35 +1701,34 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 	       u32     invalidate_domains,
 	       u32     flush_domains)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_FLUSH);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_FLUSH);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 	return 0;
 }
 
 static int
 i9xx_add_request(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_STORE_DWORD_INDEX);
-	intel_ring_emit(engine,
-			I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-	intel_ring_emit(engine, req->fence.seqno);
-	intel_ring_emit(engine, MI_USER_INTERRUPT);
-	__intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
+	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
+	intel_ring_emit(ring, req->fence.seqno);
+	intel_ring_emit(ring, MI_USER_INTERRUPT);
+	__intel_engine_submit(req->engine);
 
 	return 0;
 }
@@ -1787,20 +1795,20 @@ i965_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 length,
 			 unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine,
+	intel_ring_emit(ring,
 			MI_BATCH_BUFFER_START |
 			MI_BATCH_GTT |
 			(dispatch_flags & I915_DISPATCH_SECURE ?
 			 0 : MI_BATCH_NON_SECURE_I965));
-	intel_ring_emit(engine, offset);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, offset);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1814,8 +1822,8 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
-	u32 cs_offset = engine->scratch.gtt_offset;
+	struct intel_ringbuffer *ring = req->ringbuf;
+	u32 cs_offset = req->engine->scratch.gtt_offset;
 	int ret;
 
 	ret = intel_ring_begin(req, 6);
@@ -1823,13 +1831,13 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 		return ret;
 
 	/* Evict the invalid PTE TLBs */
-	intel_ring_emit(engine, COLOR_BLT_CMD | BLT_WRITE_RGBA);
-	intel_ring_emit(engine, BLT_DEPTH_32 | BLT_ROP_COLOR_COPY | 4096);
-	intel_ring_emit(engine, I830_TLB_ENTRIES << 16 | 4); /* load each page */
-	intel_ring_emit(engine, cs_offset);
-	intel_ring_emit(engine, 0xdeadbeef);
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, COLOR_BLT_CMD | BLT_WRITE_RGBA);
+	intel_ring_emit(ring, BLT_DEPTH_32 | BLT_ROP_COLOR_COPY | 4096);
+	intel_ring_emit(ring, I830_TLB_ENTRIES << 16 | 4); /* load each page */
+	intel_ring_emit(ring, cs_offset);
+	intel_ring_emit(ring, 0xdeadbeef);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	if ((dispatch_flags & I915_DISPATCH_PINNED) == 0) {
 		if (len > I830_BATCH_LIMIT)
@@ -1843,17 +1851,17 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 		 * stable batch scratch bo area (so that the CS never
 		 * stumbles over its tlb invalidation bug) ...
 		 */
-		intel_ring_emit(engine, SRC_COPY_BLT_CMD | BLT_WRITE_RGBA);
-		intel_ring_emit(engine,
+		intel_ring_emit(ring, SRC_COPY_BLT_CMD | BLT_WRITE_RGBA);
+		intel_ring_emit(ring,
 				BLT_DEPTH_32 | BLT_ROP_SRC_COPY | 4096);
-		intel_ring_emit(engine, DIV_ROUND_UP(len, 4096) << 16 | 4096);
-		intel_ring_emit(engine, cs_offset);
-		intel_ring_emit(engine, 4096);
-		intel_ring_emit(engine, offset);
+		intel_ring_emit(ring, DIV_ROUND_UP(len, 4096) << 16 | 4096);
+		intel_ring_emit(ring, cs_offset);
+		intel_ring_emit(ring, 4096);
+		intel_ring_emit(ring, offset);
 
-		intel_ring_emit(engine, MI_FLUSH);
-		intel_ring_emit(engine, MI_NOOP);
-		intel_ring_advance(engine);
+		intel_ring_emit(ring, MI_FLUSH);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_advance(ring);
 
 		/* ... and execute it. */
 		offset = cs_offset;
@@ -1863,10 +1871,10 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_BATCH_BUFFER_START | MI_BATCH_GTT);
-	intel_ring_emit(engine, offset | (dispatch_flags & I915_DISPATCH_SECURE ?
-					  0 : MI_BATCH_NON_SECURE));
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_BATCH_BUFFER_START | MI_BATCH_GTT);
+	intel_ring_emit(ring, offset | (dispatch_flags & I915_DISPATCH_SECURE ?
+					0 : MI_BATCH_NON_SECURE));
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1876,17 +1884,17 @@ i915_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine, MI_BATCH_BUFFER_START | MI_BATCH_GTT);
-	intel_ring_emit(engine, offset | (dispatch_flags & I915_DISPATCH_SECURE ?
-					  0 : MI_BATCH_NON_SECURE));
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, MI_BATCH_BUFFER_START | MI_BATCH_GTT);
+	intel_ring_emit(ring, offset | (dispatch_flags & I915_DISPATCH_SECURE ?
+					0 : MI_BATCH_NON_SECURE));
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2418,8 +2426,9 @@ int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 /* Align the ring tail to a cacheline boundary */
 int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
-	int num_dwords = (engine->buffer->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
+	struct intel_ringbuffer *ring = req->ringbuf;
+	int num_dwords =
+		(ring->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
 	int ret;
 
 	if (num_dwords == 0)
@@ -2431,9 +2440,9 @@ int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 		return ret;
 
 	while (num_dwords--)
-		intel_ring_emit(engine, MI_NOOP);
+		intel_ring_emit(ring, MI_NOOP);
 
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2524,7 +2533,7 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 			       u32 invalidate, u32 flush)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	uint32_t cmd;
 	int ret;
 
@@ -2552,17 +2561,16 @@ static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 	if (invalidate & I915_GEM_GPU_DOMAINS)
 		cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD;
 
-	intel_ring_emit(engine, cmd);
-	intel_ring_emit(engine,
-			I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT);
+	intel_ring_emit(ring, cmd);
+	intel_ring_emit(ring, I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT);
 	if (INTEL_GEN(req->i915) >= 8) {
-		intel_ring_emit(engine, 0); /* upper addr */
-		intel_ring_emit(engine, 0); /* value */
+		intel_ring_emit(ring, 0); /* upper addr */
+		intel_ring_emit(ring, 0); /* value */
 	} else  {
-		intel_ring_emit(engine, 0);
-		intel_ring_emit(engine, MI_NOOP);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, MI_NOOP);
 	}
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 	return 0;
 }
 
@@ -2571,8 +2579,8 @@ gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
-	bool ppgtt = USES_PPGTT(engine->dev) &&
+	struct intel_ringbuffer *ring = req->ringbuf;
+	bool ppgtt = USES_PPGTT(req->i915) &&
 			!(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
 
@@ -2581,13 +2589,13 @@ gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 		return ret;
 
 	/* FIXME(BDW): Address space and security selectors. */
-	intel_ring_emit(engine, MI_BATCH_BUFFER_START_GEN8 | (ppgtt<<8) |
+	intel_ring_emit(ring, MI_BATCH_BUFFER_START_GEN8 | (ppgtt<<8) |
 			(dispatch_flags & I915_DISPATCH_RS ?
 			 MI_BATCH_RESOURCE_STREAMER : 0));
-	intel_ring_emit(engine, lower_32_bits(offset));
-	intel_ring_emit(engine, upper_32_bits(offset));
-	intel_ring_emit(engine, MI_NOOP);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, lower_32_bits(offset));
+	intel_ring_emit(ring, upper_32_bits(offset));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2597,22 +2605,22 @@ hsw_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			     u64 offset, u32 len,
 			     unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine,
+	intel_ring_emit(ring,
 			MI_BATCH_BUFFER_START |
 			(dispatch_flags & I915_DISPATCH_SECURE ?
 			 0 : MI_BATCH_PPGTT_HSW | MI_BATCH_NON_SECURE_HSW) |
 			(dispatch_flags & I915_DISPATCH_RS ?
 			 MI_BATCH_RESOURCE_STREAMER : 0));
 	/* bit0-7 is the length on GEN6+ */
-	intel_ring_emit(engine, offset);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, offset);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2622,20 +2630,20 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(engine,
+	intel_ring_emit(ring,
 			MI_BATCH_BUFFER_START |
 			(dispatch_flags & I915_DISPATCH_SECURE ?
 			 0 : MI_BATCH_NON_SECURE_I965));
 	/* bit0-7 is the length on GEN6+ */
-	intel_ring_emit(engine, offset);
-	intel_ring_advance(engine);
+	intel_ring_emit(ring, offset);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2645,7 +2653,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 static int gen6_ring_flush(struct drm_i915_gem_request *req,
 			   u32 invalidate, u32 flush)
 {
-	struct intel_engine_cs *engine = req->engine;
+	struct intel_ringbuffer *ring = req->ringbuf;
 	uint32_t cmd;
 	int ret;
 
@@ -2672,17 +2680,17 @@ static int gen6_ring_flush(struct drm_i915_gem_request *req,
 	 */
 	if (invalidate & I915_GEM_DOMAIN_RENDER)
 		cmd |= MI_INVALIDATE_TLB;
-	intel_ring_emit(engine, cmd);
-	intel_ring_emit(engine,
+	intel_ring_emit(ring, cmd);
+	intel_ring_emit(ring,
 			I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT);
 	if (INTEL_GEN(req->i915) >= 8) {
-		intel_ring_emit(engine, 0); /* upper addr */
-		intel_ring_emit(engine, 0); /* value */
+		intel_ring_emit(ring, 0); /* upper addr */
+		intel_ring_emit(ring, 0); /* value */
 	} else  {
-		intel_ring_emit(engine, 0);
-		intel_ring_emit(engine, MI_NOOP);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring, MI_NOOP);
 	}
-	intel_ring_advance(engine);
+	intel_ring_advance(ring);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 0a830c116886..1f8e49132c54 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -454,32 +454,21 @@ int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request);
 int __must_check intel_ring_begin(struct drm_i915_gem_request *req, int n);
 int __must_check intel_ring_cacheline_align(struct drm_i915_gem_request *req);
 
-static inline void __intel_ringbuffer_emit(struct intel_ringbuffer *rb,
-					   u32 data)
+static inline void intel_ring_emit(struct intel_ringbuffer *ring, u32 data)
 {
-	*(uint32_t *)(rb->vaddr + rb->tail) = data;
-	rb->tail += 4;
+	*(uint32_t *)(ring->vaddr + ring->tail) = data;
+	ring->tail += 4;
 }
 
-static inline void __intel_ringbuffer_advance(struct intel_ringbuffer *rb)
-{
-	rb->tail &= rb->size - 1;
-}
-
-static inline void intel_ring_emit(struct intel_engine_cs *engine, u32 data)
-{
-	__intel_ringbuffer_emit(engine->buffer, data);
-}
-
-static inline void intel_ring_emit_reg(struct intel_engine_cs *engine,
+static inline void intel_ring_emit_reg(struct intel_ringbuffer *ring,
 				       i915_reg_t reg)
 {
-	intel_ring_emit(engine, i915_mmio_reg_offset(reg));
+	intel_ring_emit(ring, i915_mmio_reg_offset(reg));
 }
 
-static inline void intel_ring_advance(struct intel_engine_cs *engine)
+static inline void intel_ring_advance(struct intel_ringbuffer *ring)
 {
-	__intel_ringbuffer_advance(engine->buffer);
+	ring->tail &= ring->size - 1;
 }
 
 int __intel_ring_space(int head, int tail, int size);
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 11/31] drm/i915: Rename request->ringbuf to request->ring

[Chris Wilson]

Now that we have disambuigated ring and engine, we can use the clearer
and more consistent name for the intel_ringbuffer pointer in the
request.

@@
struct drm_i915_gem_request *r;
@@
- r->ringbuf
+ r->ring

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_context.c    |  4 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  4 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c        |  6 +-
 drivers/gpu/drm/i915/i915_gem_request.c    | 16 +++---
 drivers/gpu/drm/i915/i915_gem_request.h    |  2 +-
 drivers/gpu/drm/i915/i915_gpu_error.c      | 20 +++----
 drivers/gpu/drm/i915/intel_display.c       | 10 ++--
 drivers/gpu/drm/i915/intel_lrc.c           | 57 +++++++++---------
 drivers/gpu/drm/i915/intel_mocs.c          | 36 ++++++------
 drivers/gpu/drm/i915/intel_overlay.c       |  8 +--
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 92 +++++++++++++++---------------
 11 files changed, 126 insertions(+), 129 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index a0e24eb5e167..f7f4a8c40afe 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -552,7 +552,7 @@ static inline int
 mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 {
 	struct drm_i915_private *dev_priv = req->i915;
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	u32 flags = hw_flags | MI_MM_SPACE_GTT;
 	const int num_rings =
@@ -655,7 +655,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 static int remap_l3(struct drm_i915_gem_request *req, int slice)
 {
 	u32 *remap_info = req->i915->l3_parity.remap_info[slice];
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int i, ret;
 
 	if (!remap_info)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 793128d9410b..a8bbfe95befd 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1173,7 +1173,7 @@ i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params)
 static int
 i915_reset_gen7_sol_offsets(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret, i;
 
 	if (!IS_GEN7(req->i915) || req->engine->id != RCS) {
@@ -1303,7 +1303,7 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 
 	if (params->engine->id == RCS &&
 	    instp_mode != dev_priv->relative_constants_mode) {
-		struct intel_ringbuffer *ring = params->request->ringbuf;
+		struct intel_ringbuffer *ring = params->request->ring;
 
 		ret = intel_ring_begin(params->request, 4);
 		if (ret)
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index b38a5311f996..46cae2a92bda 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -669,7 +669,7 @@ static int gen8_write_pdp(struct drm_i915_gem_request *req,
 			  unsigned entry,
 			  dma_addr_t addr)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1661,7 +1661,7 @@ static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			 struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1688,7 +1688,7 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			  struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 49396b895a36..d2133c41be13 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -170,7 +170,7 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 	 * Note this requires that we are always called in request
 	 * completion order.
 	 */
-	request->ringbuf->last_retired_head = request->postfix;
+	request->ring->last_retired_head = request->postfix;
 
 	i915_gem_request_remove_from_client(request);
 
@@ -423,7 +423,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 			bool flush_caches)
 {
 	struct intel_engine_cs *engine;
-	struct intel_ringbuffer *ringbuf;
+	struct intel_ringbuffer *ring;
 	u32 request_start;
 	u32 reserved_tail;
 	int ret;
@@ -432,14 +432,14 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 		return;
 
 	engine = request->engine;
-	ringbuf = request->ringbuf;
+	ring = request->ring;
 
 	/*
 	 * To ensure that this call will not fail, space for its emissions
 	 * should already have been reserved in the ring buffer. Let the ring
 	 * know that it is time to use that space up.
 	 */
-	request_start = intel_ring_get_tail(ringbuf);
+	request_start = intel_ring_get_tail(ring);
 	reserved_tail = request->reserved_space;
 	request->reserved_space = 0;
 
@@ -486,21 +486,21 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 * GPU processing the request, we never over-estimate the
 	 * position of the head.
 	 */
-	request->postfix = intel_ring_get_tail(ringbuf);
+	request->postfix = intel_ring_get_tail(ring);
 
 	if (i915.enable_execlists) {
 		ret = engine->emit_request(request);
 	} else {
 		ret = engine->add_request(request);
 
-		request->tail = intel_ring_get_tail(ringbuf);
+		request->tail = intel_ring_get_tail(ring);
 	}
 	/* Not allowed to fail! */
 	WARN(ret, "emit|add_request failed: %d!\n", ret);
 	/* Sanity check that the reserved size was large enough. */
-	ret = intel_ring_get_tail(ringbuf) - request_start;
+	ret = intel_ring_get_tail(ring) - request_start;
 	if (ret < 0)
-		ret += ringbuf->size;
+		ret += ring->size;
 	WARN_ONCE(ret > reserved_tail,
 		  "Not enough space reserved (%d bytes) "
 		  "for adding the request (%d bytes)\n",
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index e06e81f459df..68868d825d9d 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -61,7 +61,7 @@ struct drm_i915_gem_request {
 	 */
 	struct i915_gem_context *ctx;
 	struct intel_engine_cs *engine;
-	struct intel_ringbuffer *ringbuf;
+	struct intel_ringbuffer *ring;
 	struct intel_signal_node signaling;
 
 	/** GEM sequence number associated with the previous request,
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 0ec917fa115d..491523c2deda 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1099,7 +1099,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 		request = i915_gem_find_active_request(engine);
 		if (request) {
 			struct i915_address_space *vm;
-			struct intel_ringbuffer *rb;
+			struct intel_ringbuffer *ring;
 
 			vm = request->ctx->ppgtt ?
 				&request->ctx->ppgtt->base : &ggtt->base;
@@ -1116,7 +1116,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 			if (HAS_BROKEN_CS_TLB(dev_priv))
 				ering->wa_batchbuffer =
 					i915_error_ggtt_object_create(dev_priv,
-							     engine->scratch.obj);
+								      engine->scratch.obj);
 
 			if (request->pid) {
 				struct task_struct *task;
@@ -1133,23 +1133,21 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 			error->simulated |=
 				request->ctx->flags & CONTEXT_NO_ERROR_CAPTURE;
 
-			rb = request->ringbuf;
-			ering->cpu_ring_head = rb->head;
-			ering->cpu_ring_tail = rb->tail;
+			ring = request->ring;
+			ering->cpu_ring_head = ring->head;
+			ering->cpu_ring_tail = ring->tail;
 			ering->ringbuffer =
 				i915_error_ggtt_object_create(dev_priv,
-							      rb->obj);
+							      ring->obj);
 		}
 
 		ering->hws_page =
 			i915_error_ggtt_object_create(dev_priv,
 						      engine->status_page.obj);
 
-		if (engine->wa_ctx.obj) {
-			ering->wa_ctx =
-				i915_error_ggtt_object_create(dev_priv,
-							      engine->wa_ctx.obj);
-		}
+		ering->wa_ctx =
+			i915_error_ggtt_object_create(dev_priv,
+						      engine->wa_ctx.obj);
 
 		i915_gem_record_active_context(engine, error, ering);
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index d18ed32e6a31..d1932840a268 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11123,7 +11123,7 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11157,7 +11157,7 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11188,7 +11188,7 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11226,7 +11226,7 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11261,7 +11261,7 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t plane_bit = 0;
 	int len, ret;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 1db2fdadbb46..15f800b75e15 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -714,7 +714,7 @@ int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request
 			return ret;
 	}
 
-	request->ringbuf = ce->ringbuf;
+	request->ring = ce->ringbuf;
 
 	if (i915.enable_guc_submission) {
 		/*
@@ -770,11 +770,11 @@ err_unpin:
 static int
 intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ringbuf = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
 
-	intel_ring_advance(ringbuf);
-	request->tail = ringbuf->tail;
+	intel_ring_advance(ring);
+	request->tail = ring->tail;
 
 	/*
 	 * Here we add two extra NOOPs as padding to avoid
@@ -782,9 +782,9 @@ intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 	 *
 	 * Caller must reserve WA_TAIL_DWORDS for us!
 	 */
-	intel_ring_emit(ringbuf, MI_NOOP);
-	intel_ring_emit(ringbuf, MI_NOOP);
-	intel_ring_advance(ringbuf);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	/* We keep the previous context alive until we retire the following
 	 * request. This ensures that any the context object is still pinned
@@ -821,7 +821,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	struct drm_device       *dev = params->dev;
 	struct intel_engine_cs *engine = params->engine;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_ringbuffer *ringbuf = params->ctx->engine[engine->id].ringbuf;
+	struct intel_ringbuffer *ring = params->request->ring;
 	u64 exec_start;
 	int instp_mode;
 	u32 instp_mask;
@@ -833,7 +833,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	case I915_EXEC_CONSTANTS_REL_GENERAL:
 	case I915_EXEC_CONSTANTS_ABSOLUTE:
 	case I915_EXEC_CONSTANTS_REL_SURFACE:
-		if (instp_mode != 0 && engine != &dev_priv->engine[RCS]) {
+		if (instp_mode != 0 && engine->id != RCS) {
 			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
 			return -EINVAL;
 		}
@@ -862,17 +862,17 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	if (ret)
 		return ret;
 
-	if (engine == &dev_priv->engine[RCS] &&
+	if (engine->id == RCS &&
 	    instp_mode != dev_priv->relative_constants_mode) {
 		ret = intel_ring_begin(params->request, 4);
 		if (ret)
 			return ret;
 
-		intel_ring_emit(ringbuf, MI_NOOP);
-		intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(1));
-		intel_ring_emit_reg(ringbuf, INSTPM);
-		intel_ring_emit(ringbuf, instp_mask << 16 | instp_mode);
-		intel_ring_advance(ringbuf);
+		intel_ring_emit(ring, MI_NOOP);
+		intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+		intel_ring_emit_reg(ring, INSTPM);
+		intel_ring_emit(ring, instp_mask << 16 | instp_mode);
+		intel_ring_advance(ring);
 
 		dev_priv->relative_constants_mode = instp_mode;
 	}
@@ -1030,7 +1030,7 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
 	int ret, i;
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct i915_workarounds *w = &req->i915->workarounds;
 
 	if (w->count == 0)
@@ -1045,14 +1045,14 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(w->count));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(w->count));
 	for (i = 0; i < w->count; i++) {
-		intel_ring_emit_reg(ringbuf, w->reg[i].addr);
-		intel_ring_emit(ringbuf, w->reg[i].value);
+		intel_ring_emit_reg(ring, w->reg[i].addr);
+		intel_ring_emit(ring, w->reg[i].value);
 	}
-	intel_ring_emit(ringbuf, MI_NOOP);
+	intel_ring_emit(ring, MI_NOOP);
 
-	intel_ring_advance(ringbuf);
+	intel_ring_advance(ring);
 
 	engine->gpu_caches_dirty = true;
 	ret = logical_ring_flush_all_caches(req);
@@ -1553,7 +1553,7 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine)
 static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 {
 	struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt;
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	const int num_lri_cmds = GEN8_LEGACY_PDPES * 2;
 	int i, ret;
@@ -1581,7 +1581,7 @@ static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
 			      u64 offset, unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	bool ppgtt = !(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
 
@@ -1638,8 +1638,7 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 			   u32 invalidate_domains,
 			   u32 unused)
 {
-	struct intel_ringbuffer *ring = request->ringbuf;
-	struct intel_engine_cs *engine = ring->engine;
+	struct intel_ringbuffer *ring = request->ring;
 	uint32_t cmd;
 	int ret;
 
@@ -1658,7 +1657,7 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 
 	if (invalidate_domains & I915_GEM_GPU_DOMAINS) {
 		cmd |= MI_INVALIDATE_TLB;
-		if (engine->id == VCS)
+		if (request->engine->id == VCS)
 			cmd |= MI_INVALIDATE_BSD;
 	}
 
@@ -1677,7 +1676,7 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 				  u32 invalidate_domains,
 				  u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
 	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	bool vf_flush_wa = false, dc_flush_wa = false;
@@ -1791,7 +1790,7 @@ static void bxt_a_seqno_barrier(struct intel_engine_cs *engine)
 
 static int gen8_emit_request(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ring = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ring;
 	int ret;
 
 	ret = intel_ring_begin(request, 6 + WA_TAIL_DWORDS);
@@ -1814,7 +1813,7 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)
 
 static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ring = request->ringbuf;
+	struct intel_ringbuffer *ring = request->ring;
 	int ret;
 
 	ret = intel_ring_begin(request, 8 + WA_TAIL_DWORDS);
diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c
index 3059c52030b4..8534ec3648bd 100644
--- a/drivers/gpu/drm/i915/intel_mocs.c
+++ b/drivers/gpu/drm/i915/intel_mocs.c
@@ -276,7 +276,7 @@ int intel_mocs_init_engine(struct intel_engine_cs *engine)
 static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 				   const struct drm_i915_mocs_table *table)
 {
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	enum intel_engine_id engine = req->engine->id;
 	unsigned int index;
 	int ret;
@@ -288,11 +288,11 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES));
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES));
 
 	for (index = 0; index < table->size; index++) {
-		intel_ring_emit_reg(ringbuf, mocs_register(engine, index));
-		intel_ring_emit(ringbuf, table->table[index].control_value);
+		intel_ring_emit_reg(ring, mocs_register(engine, index));
+		intel_ring_emit(ring, table->table[index].control_value);
 	}
 
 	/*
@@ -304,12 +304,12 @@ static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 	 * that value to all the used entries.
 	 */
 	for (; index < GEN9_NUM_MOCS_ENTRIES; index++) {
-		intel_ring_emit_reg(ringbuf, mocs_register(engine, index));
-		intel_ring_emit(ringbuf, table->table[0].control_value);
+		intel_ring_emit_reg(ring, mocs_register(engine, index));
+		intel_ring_emit(ring, table->table[0].control_value);
 	}
 
-	intel_ring_emit(ringbuf, MI_NOOP);
-	intel_ring_advance(ringbuf);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -336,7 +336,7 @@ static inline u32 l3cc_combine(const struct drm_i915_mocs_table *table,
 static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 				const struct drm_i915_mocs_table *table)
 {
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	unsigned int i;
 	int ret;
 
@@ -347,18 +347,18 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 	if (ret)
 		return ret;
 
-	intel_ring_emit(ringbuf,
+	intel_ring_emit(ring,
 			MI_LOAD_REGISTER_IMM(GEN9_NUM_MOCS_ENTRIES / 2));
 
 	for (i = 0; i < table->size/2; i++) {
-		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_ring_emit(ringbuf, l3cc_combine(table, 2*i, 2*i+1));
+		intel_ring_emit_reg(ring, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ring, l3cc_combine(table, 2*i, 2*i+1));
 	}
 
 	if (table->size & 0x01) {
 		/* Odd table size - 1 left over */
-		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_ring_emit(ringbuf, l3cc_combine(table, 2*i, 0));
+		intel_ring_emit_reg(ring, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ring, l3cc_combine(table, 2*i, 0));
 		i++;
 	}
 
@@ -368,12 +368,12 @@ static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 	 * they are reserved by the hardware.
 	 */
 	for (; i < GEN9_NUM_MOCS_ENTRIES / 2; i++) {
-		intel_ring_emit_reg(ringbuf, GEN9_LNCFCMOCS(i));
-		intel_ring_emit(ringbuf, l3cc_combine(table, 0, 0));
+		intel_ring_emit_reg(ring, GEN9_LNCFCMOCS(i));
+		intel_ring_emit(ring, l3cc_combine(table, 0, 0));
 	}
 
-	intel_ring_emit(ringbuf, MI_NOOP);
-	intel_ring_advance(ringbuf);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index ec63b64fb202..e750c0e3a267 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -253,7 +253,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 
 	overlay->active = true;
 
-	ring = req->ringbuf;
+	ring = req->ring;
 	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_ON);
 	intel_ring_emit(ring, overlay->flip_addr | OFC_UPDATE);
 	intel_ring_emit(ring, MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
@@ -295,7 +295,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 		return ret;
 	}
 
-	ring = req->ringbuf;
+	ring = req->ring;
 	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
 	intel_ring_emit(ring, flip_addr);
 	intel_ring_advance(ring);
@@ -362,7 +362,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 		return ret;
 	}
 
-	ring = req->ringbuf;
+	ring = req->ring;
 	/* wait for overlay to go idle */
 	intel_ring_emit(ring, MI_OVERLAY_FLIP | MI_OVERLAY_CONTINUE);
 	intel_ring_emit(ring, flip_addr);
@@ -438,7 +438,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 			return ret;
 		}
 
-		ring = req->ringbuf;
+		ring = req->ring;
 		intel_ring_emit(ring,
 				MI_WAIT_FOR_EVENT | MI_WAIT_FOR_OVERLAY_FLIP);
 		intel_ring_emit(ring, MI_NOOP);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 48cdfbec1f46..7ba31724feaf 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -70,7 +70,7 @@ gen2_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 cmd;
 	int ret;
 
@@ -97,7 +97,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 cmd;
 	int ret;
 
@@ -187,7 +187,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 intel_emit_post_sync_nonzero_flush(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
@@ -224,7 +224,7 @@ static int
 gen6_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
@@ -277,7 +277,7 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 gen7_render_ring_cs_stall_wa(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
@@ -299,7 +299,7 @@ static int
 gen7_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
@@ -364,7 +364,7 @@ static int
 gen8_emit_pipe_control(struct drm_i915_gem_request *req,
 		       u32 flags, u32 scratch_addr)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 6);
@@ -680,7 +680,7 @@ err:
 
 static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct i915_workarounds *w = &req->i915->workarounds;
 	int ret, i;
 
@@ -1338,7 +1338,7 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 8
-	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
+	struct intel_ringbuffer *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1380,7 +1380,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 6
-	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
+	struct intel_ringbuffer *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1419,7 +1419,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 		       unsigned int num_dwords)
 {
-	struct intel_ringbuffer *signaller = signaller_req->ringbuf;
+	struct intel_ringbuffer *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *useless;
 	enum intel_engine_id id;
@@ -1464,7 +1464,7 @@ static int
 gen6_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1488,7 +1488,7 @@ static int
 gen8_render_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1533,7 +1533,7 @@ gen8_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_ringbuffer *waiter = waiter_req->ringbuf;
+	struct intel_ringbuffer *waiter = waiter_req->ring;
 	struct drm_i915_private *dev_priv = waiter_req->i915;
 	u64 offset = GEN8_WAIT_OFFSET(waiter_req->engine, signaller->id);
 	struct i915_hw_ppgtt *ppgtt;
@@ -1567,7 +1567,7 @@ gen6_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_ringbuffer *waiter = waiter_req->ringbuf;
+	struct intel_ringbuffer *waiter = waiter_req->ring;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
@@ -1701,7 +1701,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 	       u32     invalidate_domains,
 	       u32     flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -1717,7 +1717,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 static int
 i9xx_add_request(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
@@ -1795,7 +1795,7 @@ i965_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 length,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -1822,7 +1822,7 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	u32 cs_offset = req->engine->scratch.gtt_offset;
 	int ret;
 
@@ -1884,7 +1884,7 @@ i915_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -2312,7 +2312,7 @@ int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request)
 	 */
 	request->reserved_space += LEGACY_REQUEST_SIZE;
 
-	request->ringbuf = request->engine->buffer;
+	request->ring = request->engine->buffer;
 
 	ret = intel_ring_begin(request, 0);
 	if (ret)
@@ -2324,12 +2324,12 @@ int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request)
 
 static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 {
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	struct drm_i915_gem_request *target;
 
-	intel_ring_update_space(ringbuf);
-	if (ringbuf->space >= bytes)
+	intel_ring_update_space(ring);
+	if (ring->space >= bytes)
 		return 0;
 
 	/*
@@ -2351,12 +2351,12 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 		 * from multiple ringbuffers. Here, we must ignore any that
 		 * aren't from the ringbuffer we're considering.
 		 */
-		if (target->ringbuf != ringbuf)
+		if (target->ring != ring)
 			continue;
 
 		/* Would completion of this request free enough space? */
-		space = __intel_ring_space(target->postfix, ringbuf->tail,
-					   ringbuf->size);
+		space = __intel_ring_space(target->postfix, ring->tail,
+					   ring->size);
 		if (space >= bytes)
 			break;
 	}
@@ -2369,9 +2369,9 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 
 int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 {
-	struct intel_ringbuffer *ringbuf = req->ringbuf;
-	int remain_actual = ringbuf->size - ringbuf->tail;
-	int remain_usable = ringbuf->effective_size - ringbuf->tail;
+	struct intel_ringbuffer *ring = req->ring;
+	int remain_actual = ring->size - ring->tail;
+	int remain_usable = ring->effective_size - ring->tail;
 	int bytes = num_dwords * sizeof(u32);
 	int total_bytes, wait_bytes;
 	bool need_wrap = false;
@@ -2398,35 +2398,35 @@ int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 		wait_bytes = total_bytes;
 	}
 
-	if (wait_bytes > ringbuf->space) {
+	if (wait_bytes > ring->space) {
 		int ret = wait_for_space(req, wait_bytes);
 		if (unlikely(ret))
 			return ret;
 
-		intel_ring_update_space(ringbuf);
-		if (unlikely(ringbuf->space < wait_bytes))
+		intel_ring_update_space(ring);
+		if (unlikely(ring->space < wait_bytes))
 			return -EAGAIN;
 	}
 
 	if (unlikely(need_wrap)) {
-		GEM_BUG_ON(remain_actual > ringbuf->space);
-		GEM_BUG_ON(ringbuf->tail + remain_actual > ringbuf->size);
+		GEM_BUG_ON(remain_actual > ring->space);
+		GEM_BUG_ON(ring->tail + remain_actual > ring->size);
 
 		/* Fill the tail with MI_NOOP */
-		memset(ringbuf->vaddr + ringbuf->tail, 0, remain_actual);
-		ringbuf->tail = 0;
-		ringbuf->space -= remain_actual;
+		memset(ring->vaddr + ring->tail, 0, remain_actual);
+		ring->tail = 0;
+		ring->space -= remain_actual;
 	}
 
-	ringbuf->space -= bytes;
-	GEM_BUG_ON(ringbuf->space < 0);
+	ring->space -= bytes;
+	GEM_BUG_ON(ring->space < 0);
 	return 0;
 }
 
 /* Align the ring tail to a cacheline boundary */
 int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int num_dwords =
 		(ring->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
 	int ret;
@@ -2533,7 +2533,7 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 			       u32 invalidate, u32 flush)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	uint32_t cmd;
 	int ret;
 
@@ -2579,7 +2579,7 @@ gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	bool ppgtt = USES_PPGTT(req->i915) &&
 			!(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
@@ -2605,7 +2605,7 @@ hsw_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			     u64 offset, u32 len,
 			     unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -2630,7 +2630,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -2653,7 +2653,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 static int gen6_ring_flush(struct drm_i915_gem_request *req,
 			   u32 invalidate, u32 flush)
 {
-	struct intel_ringbuffer *ring = req->ringbuf;
+	struct intel_ringbuffer *ring = req->ring;
 	uint32_t cmd;
 	int ret;
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 12/31] drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs

[Chris Wilson]

Having ringbuf->ring point to an engine is confusing, so rename it once
again to ring->engine.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 7ba31724feaf..c4582b9a6862 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -2186,7 +2186,7 @@ static void intel_ring_context_unpin(struct i915_gem_context *ctx,
 	i915_gem_context_put(ctx);
 }
 
-static int intel_init_ring_buffer(struct intel_engine_cs *engine)
+static int intel_init_engine(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	struct intel_ringbuffer *ringbuf;
@@ -2883,7 +2883,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
 	engine->init_hw = init_render_ring;
 	engine->cleanup = render_ring_cleanup;
 
-	ret = intel_init_ring_buffer(engine);
+	ret = intel_init_engine(engine);
 	if (ret)
 		return ret;
 
@@ -2922,7 +2922,7 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 			engine->irq_enable_mask = I915_BSD_USER_INTERRUPT;
 	}
 
-	return intel_init_ring_buffer(engine);
+	return intel_init_engine(engine);
 }
 
 /**
@@ -2936,7 +2936,7 @@ int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine)
 
 	engine->flush = gen6_bsd_ring_flush;
 
-	return intel_init_ring_buffer(engine);
+	return intel_init_engine(engine);
 }
 
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
@@ -2949,7 +2949,7 @@ int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
 	if (INTEL_GEN(dev_priv) < 8)
 		engine->irq_enable_mask = GT_BLT_USER_INTERRUPT;
 
-	return intel_init_ring_buffer(engine);
+	return intel_init_engine(engine);
 }
 
 int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
@@ -2966,7 +2966,7 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
 		engine->irq_disable = hsw_vebox_irq_disable;
 	}
 
-	return intel_init_ring_buffer(engine);
+	return intel_init_engine(engine);
 }
 
 int
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 13/31] drm/i915: Rename intel_context[engine].ringbuf

[Chris Wilson]

Perform s/ringbuf/ring/ on the context struct for consistency with the
ring/engine split.

v2: Kill an outdated error_ringbuf label

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_debugfs.c        |  8 +++----
 drivers/gpu/drm/i915/i915_drv.h            |  2 +-
 drivers/gpu/drm/i915/i915_gem_context.c    |  4 ++--
 drivers/gpu/drm/i915/i915_guc_submission.c |  2 +-
 drivers/gpu/drm/i915/intel_lrc.c           | 37 ++++++++++++++----------------
 5 files changed, 25 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 9aa62c5b5f65..bde68741809b 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -425,8 +425,8 @@ static int per_file_ctx_stats(int id, void *ptr, void *data)
 	for (n = 0; n < ARRAY_SIZE(ctx->engine); n++) {
 		if (ctx->engine[n].state)
 			per_file_stats(0, ctx->engine[n].state, data);
-		if (ctx->engine[n].ringbuf)
-			per_file_stats(0, ctx->engine[n].ringbuf->obj, data);
+		if (ctx->engine[n].ring)
+			per_file_stats(0, ctx->engine[n].ring->obj, data);
 	}
 
 	return 0;
@@ -2066,8 +2066,8 @@ static int i915_context_status(struct seq_file *m, void *unused)
 			seq_putc(m, ce->initialised ? 'I' : 'i');
 			if (ce->state)
 				describe_obj(m, ce->state);
-			if (ce->ringbuf)
-				describe_ctx_ringbuf(m, ce->ringbuf);
+			if (ce->ring)
+				describe_ctx_ringbuf(m, ce->ring);
 			seq_putc(m, '\n');
 		}
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 7e24ea1d465c..ecf20779e846 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -894,7 +894,7 @@ struct i915_gem_context {
 
 	struct intel_context {
 		struct drm_i915_gem_object *state;
-		struct intel_ringbuffer *ringbuf;
+		struct intel_ringbuffer *ring;
 		struct i915_vma *lrc_vma;
 		uint32_t *lrc_reg_state;
 		u64 lrc_desc;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index f7f4a8c40afe..f825b1e4aadf 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -173,8 +173,8 @@ void i915_gem_context_free(struct kref *ctx_ref)
 			continue;
 
 		WARN_ON(ce->pin_count);
-		if (ce->ringbuf)
-			intel_ringbuffer_free(ce->ringbuf);
+		if (ce->ring)
+			intel_ringbuffer_free(ce->ring);
 
 		i915_gem_object_put(ce->state);
 	}
diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index 01c1c1671811..eccd34832fe6 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -363,7 +363,7 @@ static void guc_init_ctx_desc(struct intel_guc *guc,
 		lrc->context_id = (client->ctx_index << GUC_ELC_CTXID_OFFSET) |
 				(engine->guc_id << GUC_ELC_ENGINE_OFFSET);
 
-		obj = ce->ringbuf->obj;
+		obj = ce->ring->obj;
 		gfx_addr = i915_gem_obj_ggtt_offset(obj);
 
 		lrc->ring_begin = gfx_addr;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 15f800b75e15..4c956a5b0200 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -482,11 +482,8 @@ static void execlists_context_unqueue(struct intel_engine_cs *engine)
 		 * resubmit the request. See gen8_emit_request() for where we
 		 * prepare the padding after the end of the request.
 		 */
-		struct intel_ringbuffer *ringbuf;
-
-		ringbuf = req0->ctx->engine[engine->id].ringbuf;
 		req0->tail += 8;
-		req0->tail &= ringbuf->size - 1;
+		req0->tail &= req0->ring->size - 1;
 	}
 
 	execlists_submit_requests(req0, req1);
@@ -714,7 +711,7 @@ int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request
 			return ret;
 	}
 
-	request->ring = ce->ringbuf;
+	request->ring = ce->ring;
 
 	if (i915.enable_guc_submission) {
 		/*
@@ -976,14 +973,14 @@ static int intel_lr_context_pin(struct i915_gem_context *ctx,
 
 	lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
 
-	ret = intel_pin_and_map_ringbuffer_obj(dev_priv, ce->ringbuf);
+	ret = intel_pin_and_map_ringbuffer_obj(dev_priv, ce->ring);
 	if (ret)
 		goto unpin_map;
 
 	ce->lrc_vma = i915_gem_obj_to_ggtt(ce->state);
 	intel_lr_context_descriptor_update(ctx, engine);
 
-	lrc_reg_state[CTX_RING_BUFFER_START+1] = ce->ringbuf->vma->node.start;
+	lrc_reg_state[CTX_RING_BUFFER_START+1] = ce->ring->vma->node.start;
 	ce->lrc_reg_state = lrc_reg_state;
 	ce->state->dirty = true;
 
@@ -1014,7 +1011,7 @@ void intel_lr_context_unpin(struct i915_gem_context *ctx,
 	if (--ce->pin_count)
 		return;
 
-	intel_unpin_ringbuffer_obj(ce->ringbuf);
+	intel_unpin_ringbuffer_obj(ce->ring);
 
 	i915_gem_object_unpin_map(ce->state);
 	i915_gem_object_ggtt_unpin(ce->state);
@@ -2346,7 +2343,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 	struct drm_i915_gem_object *ctx_obj;
 	struct intel_context *ce = &ctx->engine[engine->id];
 	uint32_t context_size;
-	struct intel_ringbuffer *ringbuf;
+	struct intel_ringbuffer *ring;
 	int ret;
 
 	WARN_ON(ce->state);
@@ -2362,29 +2359,29 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 		return PTR_ERR(ctx_obj);
 	}
 
-	ringbuf = intel_engine_create_ringbuffer(engine, ctx->ring_size);
-	if (IS_ERR(ringbuf)) {
-		ret = PTR_ERR(ringbuf);
+	ring = intel_engine_create_ringbuffer(engine, ctx->ring_size);
+	if (IS_ERR(ring)) {
+		ret = PTR_ERR(ring);
 		goto error_deref_obj;
 	}
 
-	ret = populate_lr_context(ctx, ctx_obj, engine, ringbuf);
+	ret = populate_lr_context(ctx, ctx_obj, engine, ring);
 	if (ret) {
 		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
-		goto error_ringbuf;
+		goto error_ring_free;
 	}
 
-	ce->ringbuf = ringbuf;
+	ce->ring = ring;
 	ce->state = ctx_obj;
 	ce->initialised = engine->init_context == NULL;
 
 	return 0;
 
-error_ringbuf:
-	intel_ringbuffer_free(ringbuf);
+error_ring_free:
+	intel_ringbuffer_free(ring);
 error_deref_obj:
 	i915_gem_object_put(ctx_obj);
-	ce->ringbuf = NULL;
+	ce->ring = NULL;
 	ce->state = NULL;
 	return ret;
 }
@@ -2415,7 +2412,7 @@ void intel_lr_context_reset(struct drm_i915_private *dev_priv,
 
 		i915_gem_object_unpin_map(ctx_obj);
 
-		ce->ringbuf->head = 0;
-		ce->ringbuf->tail = 0;
+		ce->ring->head = 0;
+		ce->ring->tail = 0;
 	}
 }
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 14/31] drm/i915: Rename struct intel_ringbuffer to struct intel_ring

[Chris Wilson]

The state stored in this struct is not only the information about the
buffer object, but the ring used to communicate with the hardware. Using
buffer here is overly specific and, for me at least, conflates with the
notion of buffer objects themselves.

s/struct intel_ringbuffer/struct intel_ring/
s/enum intel_ring_hangcheck/enum intel_engine_hangcheck/
s/describe_ctx_ringbuf()/describe_ctx_ring()/
s/intel_ring_get_active_head()/intel_engine_get_active_head()/
s/intel_ring_sync_index()/intel_engine_sync_index()/
s/intel_ring_init_seqno()/intel_engine_init_seqno()/
s/ring_stuck()/engine_stuck()/
s/intel_cleanup_engine()/intel_engine_cleanup()/
s/intel_stop_engine()/intel_engine_stop()/
s/intel_pin_and_map_ringbuffer_obj()/intel_pin_and_map_ring()/
s/intel_unpin_ringbuffer()/intel_unpin_ring()/
s/intel_engine_create_ringbuffer()/intel_engine_create_ring()/
s/intel_ring_flush_all_caches()/intel_engine_flush_all_caches()/
s/intel_ring_invalidate_all_caches()/intel_engine_invalidate_all_caches()/
s/intel_ringbuffer_free()/intel_ring_free()/

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_debugfs.c        |  11 ++-
 drivers/gpu/drm/i915/i915_drv.h            |   4 +-
 drivers/gpu/drm/i915/i915_gem.c            |  16 ++--
 drivers/gpu/drm/i915/i915_gem_context.c    |   6 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |   6 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c        |   6 +-
 drivers/gpu/drm/i915/i915_gem_request.c    |   6 +-
 drivers/gpu/drm/i915/i915_gem_request.h    |   2 +-
 drivers/gpu/drm/i915/i915_gpu_error.c      |   8 +-
 drivers/gpu/drm/i915/i915_irq.c            |  14 ++--
 drivers/gpu/drm/i915/intel_display.c       |  10 +--
 drivers/gpu/drm/i915/intel_engine_cs.c     |   2 +-
 drivers/gpu/drm/i915/intel_lrc.c           |  34 ++++----
 drivers/gpu/drm/i915/intel_mocs.c          |   4 +-
 drivers/gpu/drm/i915/intel_overlay.c       |   8 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 128 ++++++++++++++---------------
 drivers/gpu/drm/i915/intel_ringbuffer.h    |  51 ++++++------
 17 files changed, 157 insertions(+), 159 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index bde68741809b..dccc72d63dd0 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -1419,7 +1419,7 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused)
 	intel_runtime_pm_get(dev_priv);
 
 	for_each_engine_id(engine, dev_priv, id) {
-		acthd[id] = intel_ring_get_active_head(engine);
+		acthd[id] = intel_engine_get_active_head(engine);
 		seqno[id] = intel_engine_get_seqno(engine);
 	}
 
@@ -2017,12 +2017,11 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 	return 0;
 }
 
-static void describe_ctx_ringbuf(struct seq_file *m,
-				 struct intel_ringbuffer *ringbuf)
+static void describe_ctx_ring(struct seq_file *m, struct intel_ring *ring)
 {
 	seq_printf(m, " (ringbuffer, space: %d, head: %u, tail: %u, last head: %d)",
-		   ringbuf->space, ringbuf->head, ringbuf->tail,
-		   ringbuf->last_retired_head);
+		   ring->space, ring->head, ring->tail,
+		   ring->last_retired_head);
 }
 
 static int i915_context_status(struct seq_file *m, void *unused)
@@ -2067,7 +2066,7 @@ static int i915_context_status(struct seq_file *m, void *unused)
 			if (ce->state)
 				describe_obj(m, ce->state);
 			if (ce->ring)
-				describe_ctx_ringbuf(m, ce->ring);
+				describe_ctx_ring(m, ce->ring);
 			seq_putc(m, '\n');
 		}
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index ecf20779e846..09e72b0510ab 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -518,7 +518,7 @@ struct drm_i915_error_state {
 		bool waiting;
 		int num_waiters;
 		int hangcheck_score;
-		enum intel_ring_hangcheck_action hangcheck_action;
+		enum intel_engine_hangcheck_action hangcheck_action;
 		int num_requests;
 
 		/* our own tracking of ring head and tail */
@@ -894,7 +894,7 @@ struct i915_gem_context {
 
 	struct intel_context {
 		struct drm_i915_gem_object *state;
-		struct intel_ringbuffer *ring;
+		struct intel_ring *ring;
 		struct i915_vma *lrc_vma;
 		uint32_t *lrc_reg_state;
 		u64 lrc_desc;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 7bfce1d5c61b..59890f523c5f 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2486,7 +2486,7 @@ static void i915_gem_reset_engine_status(struct intel_engine_cs *engine)
 
 static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
 {
-	struct intel_ringbuffer *buffer;
+	struct intel_ring *ring;
 
 	while (!list_empty(&engine->active_list)) {
 		struct drm_i915_gem_object *obj;
@@ -2502,7 +2502,7 @@ static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
 	 * (lockless) lookup doesn't try and wait upon the request as we
 	 * reset it.
 	 */
-	intel_ring_init_seqno(engine, engine->last_submitted_seqno);
+	intel_engine_init_seqno(engine, engine->last_submitted_seqno);
 
 	/*
 	 * Clear the execlists queue up before freeing the requests, as those
@@ -2541,9 +2541,9 @@ static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
 	 * upon reset is less than when we start. Do one more pass over
 	 * all the ringbuffers to reset last_retired_head.
 	 */
-	list_for_each_entry(buffer, &engine->buffers, link) {
-		buffer->last_retired_head = buffer->tail;
-		intel_ring_update_space(buffer);
+	list_for_each_entry(ring, &engine->buffers, link) {
+		ring->last_retired_head = ring->tail;
+		intel_ring_update_space(ring);
 	}
 
 	engine->i915->gt.active_engines &= ~intel_engine_flag(engine);
@@ -2870,7 +2870,7 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
 
 		i915_gem_object_retire_request(obj, from_req);
 	} else {
-		int idx = intel_ring_sync_index(from, to);
+		int idx = intel_engine_sync_index(from, to);
 		u32 seqno = i915_gem_request_get_seqno(from_req);
 
 		WARN_ON(!to_req);
@@ -4570,8 +4570,8 @@ int i915_gem_init(struct drm_device *dev)
 
 	if (!i915.enable_execlists) {
 		dev_priv->gt.execbuf_submit = i915_gem_ringbuffer_submission;
-		dev_priv->gt.cleanup_engine = intel_cleanup_engine;
-		dev_priv->gt.stop_engine = intel_stop_engine;
+		dev_priv->gt.cleanup_engine = intel_engine_cleanup;
+		dev_priv->gt.stop_engine = intel_engine_stop;
 	} else {
 		dev_priv->gt.execbuf_submit = intel_execlists_submission;
 		dev_priv->gt.cleanup_engine = intel_logical_ring_cleanup;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index f825b1e4aadf..3336a5fcd029 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -174,7 +174,7 @@ void i915_gem_context_free(struct kref *ctx_ref)
 
 		WARN_ON(ce->pin_count);
 		if (ce->ring)
-			intel_ringbuffer_free(ce->ring);
+			intel_ring_free(ce->ring);
 
 		i915_gem_object_put(ce->state);
 	}
@@ -552,7 +552,7 @@ static inline int
 mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 {
 	struct drm_i915_private *dev_priv = req->i915;
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	u32 flags = hw_flags | MI_MM_SPACE_GTT;
 	const int num_rings =
@@ -655,7 +655,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 static int remap_l3(struct drm_i915_gem_request *req, int slice)
 {
 	u32 *remap_info = req->i915->l3_parity.remap_info[slice];
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int i, ret;
 
 	if (!remap_info)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a8bbfe95befd..ca1d4f573832 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1001,7 +1001,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 	/* Unconditionally invalidate gpu caches and ensure that we do flush
 	 * any residual writes from the previous batch.
 	 */
-	return intel_ring_invalidate_all_caches(req);
+	return intel_engine_invalidate_all_caches(req);
 }
 
 static bool
@@ -1173,7 +1173,7 @@ i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params)
 static int
 i915_reset_gen7_sol_offsets(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret, i;
 
 	if (!IS_GEN7(req->i915) || req->engine->id != RCS) {
@@ -1303,7 +1303,7 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 
 	if (params->engine->id == RCS &&
 	    instp_mode != dev_priv->relative_constants_mode) {
-		struct intel_ringbuffer *ring = params->request->ring;
+		struct intel_ring *ring = params->request->ring;
 
 		ret = intel_ring_begin(params->request, 4);
 		if (ret)
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 46cae2a92bda..ebfa0406a6a1 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -669,7 +669,7 @@ static int gen8_write_pdp(struct drm_i915_gem_request *req,
 			  unsigned entry,
 			  dma_addr_t addr)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1661,7 +1661,7 @@ static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			 struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
@@ -1688,7 +1688,7 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 			  struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index d2133c41be13..942b5b1f1602 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -244,7 +244,7 @@ static int i915_gem_init_seqno(struct drm_i915_private *dev_priv, u32 seqno)
 
 	/* Finally reset hw state */
 	for_each_engine(engine, dev_priv)
-		intel_ring_init_seqno(engine, seqno);
+		intel_engine_init_seqno(engine, seqno);
 
 	return 0;
 }
@@ -423,7 +423,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 			bool flush_caches)
 {
 	struct intel_engine_cs *engine;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	u32 request_start;
 	u32 reserved_tail;
 	int ret;
@@ -454,7 +454,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 		if (i915.enable_execlists)
 			ret = logical_ring_flush_all_caches(request);
 		else
-			ret = intel_ring_flush_all_caches(request);
+			ret = intel_engine_flush_all_caches(request);
 		/* Not allowed to fail! */
 		WARN(ret, "*_ring_flush_all_caches failed: %d!\n", ret);
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index 68868d825d9d..382ca5a163eb 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -61,7 +61,7 @@ struct drm_i915_gem_request {
 	 */
 	struct i915_gem_context *ctx;
 	struct intel_engine_cs *engine;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	struct intel_signal_node signaling;
 
 	/** GEM sequence number associated with the previous request,
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 491523c2deda..3d40b1dffc73 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -221,7 +221,7 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 	}
 }
 
-static const char *hangcheck_action_to_str(enum intel_ring_hangcheck_action a)
+static const char *hangcheck_action_to_str(enum intel_engine_hangcheck_action a)
 {
 	switch (a) {
 	case HANGCHECK_IDLE:
@@ -889,7 +889,7 @@ static void gen8_record_semaphore_state(struct drm_i915_private *dev_priv,
 		signal_offset = (GEN8_SIGNAL_OFFSET(engine, id) & (PAGE_SIZE - 1))
 				/ 4;
 		tmp = error->semaphore_obj->pages[0];
-		idx = intel_ring_sync_index(engine, to);
+		idx = intel_engine_sync_index(engine, to);
 
 		ering->semaphore_mboxes[idx] = tmp[signal_offset];
 		ering->semaphore_seqno[idx] = engine->semaphore.sync_seqno[idx];
@@ -990,7 +990,7 @@ static void i915_record_ring_state(struct drm_i915_private *dev_priv,
 
 	ering->waiting = intel_engine_has_waiter(engine);
 	ering->instpm = I915_READ(RING_INSTPM(engine->mmio_base));
-	ering->acthd = intel_ring_get_active_head(engine);
+	ering->acthd = intel_engine_get_active_head(engine);
 	ering->seqno = intel_engine_get_seqno(engine);
 	ering->last_seqno = engine->last_submitted_seqno;
 	ering->start = I915_READ_START(engine);
@@ -1099,7 +1099,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 		request = i915_gem_find_active_request(engine);
 		if (request) {
 			struct i915_address_space *vm;
-			struct intel_ringbuffer *ring;
+			struct intel_ring *ring;
 
 			vm = request->ctx->ppgtt ?
 				&request->ctx->ppgtt->base : &ggtt->base;
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index f5bf4f913a91..e58650096426 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -2993,7 +2993,7 @@ static bool subunits_stuck(struct intel_engine_cs *engine)
 	return stuck;
 }
 
-static enum intel_ring_hangcheck_action
+static enum intel_engine_hangcheck_action
 head_stuck(struct intel_engine_cs *engine, u64 acthd)
 {
 	if (acthd != engine->hangcheck.acthd) {
@@ -3011,11 +3011,11 @@ head_stuck(struct intel_engine_cs *engine, u64 acthd)
 	return HANGCHECK_HUNG;
 }
 
-static enum intel_ring_hangcheck_action
-ring_stuck(struct intel_engine_cs *engine, u64 acthd)
+static enum intel_engine_hangcheck_action
+engine_stuck(struct intel_engine_cs *engine, u64 acthd)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	enum intel_ring_hangcheck_action ha;
+	enum intel_engine_hangcheck_action ha;
 	u32 tmp;
 
 	ha = head_stuck(engine, acthd);
@@ -3124,7 +3124,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 		if (engine->irq_seqno_barrier)
 			engine->irq_seqno_barrier(engine);
 
-		acthd = intel_ring_get_active_head(engine);
+		acthd = intel_engine_get_active_head(engine);
 		seqno = intel_engine_get_seqno(engine);
 
 		/* Reset stuck interrupts between batch advances */
@@ -3154,8 +3154,8 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 				 * being repeatedly kicked and so responsible
 				 * for stalling the machine.
 				 */
-				engine->hangcheck.action = ring_stuck(engine,
-								      acthd);
+				engine->hangcheck.action =
+					engine_stuck(engine, acthd);
 
 				switch (engine->hangcheck.action) {
 				case HANGCHECK_IDLE:
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index d1932840a268..bff172c45ff7 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11123,7 +11123,7 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11157,7 +11157,7 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
 	int ret;
@@ -11188,7 +11188,7 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11226,7 +11226,7 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
@@ -11261,7 +11261,7 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 				 struct drm_i915_gem_request *req,
 				 uint32_t flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t plane_bit = 0;
 	int len, ret;
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index f4a35ec78481..f00bd55fe582 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -154,7 +154,7 @@ cleanup:
 		if (i915.enable_execlists)
 			intel_logical_ring_cleanup(&dev_priv->engine[i]);
 		else
-			intel_cleanup_engine(&dev_priv->engine[i]);
+			intel_engine_cleanup(&dev_priv->engine[i]);
 	}
 
 	return ret;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 4c956a5b0200..beb82ef1c321 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -767,7 +767,7 @@ err_unpin:
 static int
 intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ring = request->ring;
+	struct intel_ring *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
 
 	intel_ring_advance(ring);
@@ -818,7 +818,7 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	struct drm_device       *dev = params->dev;
 	struct intel_engine_cs *engine = params->engine;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_ringbuffer *ring = params->request->ring;
+	struct intel_ring *ring = params->request->ring;
 	u64 exec_start;
 	int instp_mode;
 	u32 instp_mask;
@@ -973,7 +973,7 @@ static int intel_lr_context_pin(struct i915_gem_context *ctx,
 
 	lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
 
-	ret = intel_pin_and_map_ringbuffer_obj(dev_priv, ce->ring);
+	ret = intel_pin_and_map_ring(dev_priv, ce->ring);
 	if (ret)
 		goto unpin_map;
 
@@ -1011,7 +1011,7 @@ void intel_lr_context_unpin(struct i915_gem_context *ctx,
 	if (--ce->pin_count)
 		return;
 
-	intel_unpin_ringbuffer_obj(ce->ring);
+	intel_unpin_ring(ce->ring);
 
 	i915_gem_object_unpin_map(ce->state);
 	i915_gem_object_ggtt_unpin(ce->state);
@@ -1027,7 +1027,7 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
 	int ret, i;
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct i915_workarounds *w = &req->i915->workarounds;
 
 	if (w->count == 0)
@@ -1550,7 +1550,7 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine)
 static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 {
 	struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt;
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	const int num_lri_cmds = GEN8_LEGACY_PDPES * 2;
 	int i, ret;
@@ -1578,7 +1578,7 @@ static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
 			      u64 offset, unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	bool ppgtt = !(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
 
@@ -1635,8 +1635,8 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 			   u32 invalidate_domains,
 			   u32 unused)
 {
-	struct intel_ringbuffer *ring = request->ring;
-	uint32_t cmd;
+	struct intel_ring *ring = request->ring;
+	u32 cmd;
 	int ret;
 
 	ret = intel_ring_begin(request, 4);
@@ -1673,7 +1673,7 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 				  u32 invalidate_domains,
 				  u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = request->ring;
+	struct intel_ring *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
 	u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	bool vf_flush_wa = false, dc_flush_wa = false;
@@ -1787,7 +1787,7 @@ static void bxt_a_seqno_barrier(struct intel_engine_cs *engine)
 
 static int gen8_emit_request(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ring = request->ring;
+	struct intel_ring *ring = request->ring;
 	int ret;
 
 	ret = intel_ring_begin(request, 6 + WA_TAIL_DWORDS);
@@ -1810,7 +1810,7 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)
 
 static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 {
-	struct intel_ringbuffer *ring = request->ring;
+	struct intel_ring *ring = request->ring;
 	int ret;
 
 	ret = intel_ring_begin(request, 8 + WA_TAIL_DWORDS);
@@ -2162,7 +2162,7 @@ static int
 populate_lr_context(struct i915_gem_context *ctx,
 		    struct drm_i915_gem_object *ctx_obj,
 		    struct intel_engine_cs *engine,
-		    struct intel_ringbuffer *ringbuf)
+		    struct intel_ring *ring)
 {
 	struct drm_i915_private *dev_priv = ctx->i915;
 	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt;
@@ -2215,7 +2215,7 @@ populate_lr_context(struct i915_gem_context *ctx,
 		       RING_START(engine->mmio_base), 0);
 	ASSIGN_CTX_REG(reg_state, CTX_RING_BUFFER_CONTROL,
 		       RING_CTL(engine->mmio_base),
-		       ((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES) | RING_VALID);
+		       ((ring->size - PAGE_SIZE) & RING_NR_PAGES) | RING_VALID);
 	ASSIGN_CTX_REG(reg_state, CTX_BB_HEAD_U,
 		       RING_BBADDR_UDW(engine->mmio_base), 0);
 	ASSIGN_CTX_REG(reg_state, CTX_BB_HEAD_L,
@@ -2343,7 +2343,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 	struct drm_i915_gem_object *ctx_obj;
 	struct intel_context *ce = &ctx->engine[engine->id];
 	uint32_t context_size;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	int ret;
 
 	WARN_ON(ce->state);
@@ -2359,7 +2359,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 		return PTR_ERR(ctx_obj);
 	}
 
-	ring = intel_engine_create_ringbuffer(engine, ctx->ring_size);
+	ring = intel_engine_create_ring(engine, ctx->ring_size);
 	if (IS_ERR(ring)) {
 		ret = PTR_ERR(ring);
 		goto error_deref_obj;
@@ -2378,7 +2378,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 	return 0;
 
 error_ring_free:
-	intel_ringbuffer_free(ring);
+	intel_ring_free(ring);
 error_deref_obj:
 	i915_gem_object_put(ctx_obj);
 	ce->ring = NULL;
diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c
index 8534ec3648bd..80bb9247ce66 100644
--- a/drivers/gpu/drm/i915/intel_mocs.c
+++ b/drivers/gpu/drm/i915/intel_mocs.c
@@ -276,7 +276,7 @@ int intel_mocs_init_engine(struct intel_engine_cs *engine)
 static int emit_mocs_control_table(struct drm_i915_gem_request *req,
 				   const struct drm_i915_mocs_table *table)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	enum intel_engine_id engine = req->engine->id;
 	unsigned int index;
 	int ret;
@@ -336,7 +336,7 @@ static inline u32 l3cc_combine(const struct drm_i915_mocs_table *table,
 static int emit_mocs_l3cc_table(struct drm_i915_gem_request *req,
 				const struct drm_i915_mocs_table *table)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	unsigned int i;
 	int ret;
 
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index e750c0e3a267..8f1d4d9ef345 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -235,7 +235,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	int ret;
 
 	WARN_ON(overlay->active);
@@ -270,7 +270,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	u32 flip_addr = overlay->flip_addr;
 	u32 tmp;
 	int ret;
@@ -340,7 +340,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 	struct drm_i915_private *dev_priv = overlay->i915;
 	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	u32 flip_addr = overlay->flip_addr;
 	int ret;
 
@@ -426,7 +426,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 	if (I915_READ(ISR) & I915_OVERLAY_PLANE_FLIP_PENDING_INTERRUPT) {
 		/* synchronous slowpath */
 		struct drm_i915_gem_request *req;
-		struct intel_ringbuffer *ring;
+		struct intel_ring *ring;
 
 		req = i915_gem_request_alloc(engine, NULL);
 		if (IS_ERR(req))
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c4582b9a6862..90572b862eec 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -47,7 +47,7 @@ int __intel_ring_space(int head, int tail, int size)
 	return space - I915_RING_FREE_SPACE;
 }
 
-void intel_ring_update_space(struct intel_ringbuffer *ringbuf)
+void intel_ring_update_space(struct intel_ring *ringbuf)
 {
 	if (ringbuf->last_retired_head != -1) {
 		ringbuf->head = ringbuf->last_retired_head;
@@ -60,9 +60,10 @@ void intel_ring_update_space(struct intel_ringbuffer *ringbuf)
 
 static void __intel_engine_submit(struct intel_engine_cs *engine)
 {
-	struct intel_ringbuffer *ringbuf = engine->buffer;
-	ringbuf->tail &= ringbuf->size - 1;
-	engine->write_tail(engine, ringbuf->tail);
+	struct intel_ring *ring = engine->buffer;
+
+	ring->tail &= ring->size - 1;
+	engine->write_tail(engine, ring->tail);
 }
 
 static int
@@ -70,7 +71,7 @@ gen2_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 cmd;
 	int ret;
 
@@ -97,7 +98,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 cmd;
 	int ret;
 
@@ -187,7 +188,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 intel_emit_post_sync_nonzero_flush(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
@@ -224,7 +225,7 @@ static int
 gen6_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
@@ -277,7 +278,7 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 static int
 gen7_render_ring_cs_stall_wa(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
@@ -299,7 +300,7 @@ static int
 gen7_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32 invalidate_domains, u32 flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 scratch_addr =
 		req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
@@ -364,7 +365,7 @@ static int
 gen8_emit_pipe_control(struct drm_i915_gem_request *req,
 		       u32 flags, u32 scratch_addr)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 6);
@@ -427,7 +428,7 @@ static void ring_write_tail(struct intel_engine_cs *engine,
 	I915_WRITE_TAIL(engine, value);
 }
 
-u64 intel_ring_get_active_head(struct intel_engine_cs *engine)
+u64 intel_engine_get_active_head(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	u64 acthd;
@@ -553,8 +554,8 @@ static bool stop_ring(struct intel_engine_cs *engine)
 static int init_ring_common(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	struct intel_ringbuffer *ringbuf = engine->buffer;
-	struct drm_i915_gem_object *obj = ringbuf->obj;
+	struct intel_ring *ring = engine->buffer;
+	struct drm_i915_gem_object *obj = ring->obj;
 	int ret = 0;
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
@@ -604,7 +605,7 @@ static int init_ring_common(struct intel_engine_cs *engine)
 	(void)I915_READ_HEAD(engine);
 
 	I915_WRITE_CTL(engine,
-			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES)
+			((ring->size - PAGE_SIZE) & RING_NR_PAGES)
 			| RING_VALID);
 
 	/* If the head is still not zero, the ring is dead */
@@ -623,10 +624,10 @@ static int init_ring_common(struct intel_engine_cs *engine)
 		goto out;
 	}
 
-	ringbuf->last_retired_head = -1;
-	ringbuf->head = I915_READ_HEAD(engine);
-	ringbuf->tail = I915_READ_TAIL(engine) & TAIL_ADDR;
-	intel_ring_update_space(ringbuf);
+	ring->last_retired_head = -1;
+	ring->head = I915_READ_HEAD(engine);
+	ring->tail = I915_READ_TAIL(engine) & TAIL_ADDR;
+	intel_ring_update_space(ring);
 
 	intel_engine_init_hangcheck(engine);
 
@@ -680,7 +681,7 @@ err:
 
 static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct i915_workarounds *w = &req->i915->workarounds;
 	int ret, i;
 
@@ -688,7 +689,7 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 		return 0;
 
 	req->engine->gpu_caches_dirty = true;
-	ret = intel_ring_flush_all_caches(req);
+	ret = intel_engine_flush_all_caches(req);
 	if (ret)
 		return ret;
 
@@ -706,7 +707,7 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	intel_ring_advance(ring);
 
 	req->engine->gpu_caches_dirty = true;
-	ret = intel_ring_flush_all_caches(req);
+	ret = intel_engine_flush_all_caches(req);
 	if (ret)
 		return ret;
 
@@ -1338,7 +1339,7 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 8
-	struct intel_ringbuffer *signaller = signaller_req->ring;
+	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1380,7 +1381,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 			   unsigned int num_dwords)
 {
 #define MBOX_UPDATE_DWORDS 6
-	struct intel_ringbuffer *signaller = signaller_req->ring;
+	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
@@ -1419,7 +1420,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 		       unsigned int num_dwords)
 {
-	struct intel_ringbuffer *signaller = signaller_req->ring;
+	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *useless;
 	enum intel_engine_id id;
@@ -1464,7 +1465,7 @@ static int
 gen6_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1488,7 +1489,7 @@ static int
 gen8_render_add_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	if (engine->semaphore.signal)
@@ -1533,7 +1534,7 @@ gen8_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_ringbuffer *waiter = waiter_req->ring;
+	struct intel_ring *waiter = waiter_req->ring;
 	struct drm_i915_private *dev_priv = waiter_req->i915;
 	u64 offset = GEN8_WAIT_OFFSET(waiter_req->engine, signaller->id);
 	struct i915_hw_ppgtt *ppgtt;
@@ -1567,7 +1568,7 @@ gen6_ring_sync(struct drm_i915_gem_request *waiter_req,
 	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	struct intel_ringbuffer *waiter = waiter_req->ring;
+	struct intel_ring *waiter = waiter_req->ring;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
@@ -1701,7 +1702,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 	       u32     invalidate_domains,
 	       u32     flush_domains)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -1717,7 +1718,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 static int
 i9xx_add_request(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 4);
@@ -1795,7 +1796,7 @@ i965_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 length,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -1822,7 +1823,7 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	u32 cs_offset = req->engine->scratch.gtt_offset;
 	int ret;
 
@@ -1884,7 +1885,7 @@ i915_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			 u64 offset, u32 len,
 			 unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -1992,7 +1993,7 @@ static int init_phys_status_page(struct intel_engine_cs *engine)
 	return 0;
 }
 
-void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
+void intel_unpin_ring(struct intel_ring *ringbuf)
 {
 	GEM_BUG_ON(!ringbuf->vma);
 	GEM_BUG_ON(!ringbuf->vaddr);
@@ -2007,8 +2008,8 @@ void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
 	ringbuf->vma = NULL;
 }
 
-int intel_pin_and_map_ringbuffer_obj(struct drm_i915_private *dev_priv,
-				     struct intel_ringbuffer *ringbuf)
+int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
+			   struct intel_ring *ringbuf)
 {
 	struct drm_i915_gem_object *obj = ringbuf->obj;
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
@@ -2060,14 +2061,14 @@ err_unpin:
 	return ret;
 }
 
-static void intel_destroy_ringbuffer_obj(struct intel_ringbuffer *ringbuf)
+static void intel_destroy_ringbuffer_obj(struct intel_ring *ringbuf)
 {
 	i915_gem_object_put(ringbuf->obj);
 	ringbuf->obj = NULL;
 }
 
 static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
-				      struct intel_ringbuffer *ringbuf)
+				      struct intel_ring *ringbuf)
 {
 	struct drm_i915_gem_object *obj;
 
@@ -2087,10 +2088,10 @@ static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
 	return 0;
 }
 
-struct intel_ringbuffer *
-intel_engine_create_ringbuffer(struct intel_engine_cs *engine, int size)
+struct intel_ring *
+intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 {
-	struct intel_ringbuffer *ring;
+	struct intel_ring *ring;
 	int ret;
 
 	ring = kzalloc(sizeof(*ring), GFP_KERNEL);
@@ -2128,7 +2129,7 @@ intel_engine_create_ringbuffer(struct intel_engine_cs *engine, int size)
 }
 
 void
-intel_ringbuffer_free(struct intel_ringbuffer *ring)
+intel_ring_free(struct intel_ring *ring)
 {
 	intel_destroy_ringbuffer_obj(ring);
 	list_del(&ring->link);
@@ -2189,7 +2190,7 @@ static void intel_ring_context_unpin(struct i915_gem_context *ctx,
 static int intel_init_engine(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	struct intel_ringbuffer *ringbuf;
+	struct intel_ring *ringbuf;
 	int ret;
 
 	WARN_ON(engine->buffer);
@@ -2214,7 +2215,7 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 	if (ret)
 		goto error;
 
-	ringbuf = intel_engine_create_ringbuffer(engine, 32 * PAGE_SIZE);
+	ringbuf = intel_engine_create_ring(engine, 32 * PAGE_SIZE);
 	if (IS_ERR(ringbuf)) {
 		ret = PTR_ERR(ringbuf);
 		goto error;
@@ -2232,7 +2233,7 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 			goto error;
 	}
 
-	ret = intel_pin_and_map_ringbuffer_obj(dev_priv, ringbuf);
+	ret = intel_pin_and_map_ring(dev_priv, ringbuf);
 	if (ret) {
 		DRM_ERROR("Failed to pin and map ringbuffer %s: %d\n",
 				engine->name, ret);
@@ -2243,11 +2244,11 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 	return 0;
 
 error:
-	intel_cleanup_engine(engine);
+	intel_engine_cleanup(engine);
 	return ret;
 }
 
-void intel_cleanup_engine(struct intel_engine_cs *engine)
+void intel_engine_cleanup(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv;
 
@@ -2257,11 +2258,11 @@ void intel_cleanup_engine(struct intel_engine_cs *engine)
 	dev_priv = engine->i915;
 
 	if (engine->buffer) {
-		intel_stop_engine(engine);
+		intel_engine_stop(engine);
 		WARN_ON(!IS_GEN2(dev_priv) && (I915_READ_MODE(engine) & MODE_IDLE) == 0);
 
-		intel_unpin_ringbuffer_obj(engine->buffer);
-		intel_ringbuffer_free(engine->buffer);
+		intel_unpin_ring(engine->buffer);
+		intel_ring_free(engine->buffer);
 		engine->buffer = NULL;
 	}
 
@@ -2324,7 +2325,7 @@ int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request)
 
 static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	struct intel_engine_cs *engine = req->engine;
 	struct drm_i915_gem_request *target;
 
@@ -2369,7 +2370,7 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 
 int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int remain_actual = ring->size - ring->tail;
 	int remain_usable = ring->effective_size - ring->tail;
 	int bytes = num_dwords * sizeof(u32);
@@ -2426,7 +2427,7 @@ int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 /* Align the ring tail to a cacheline boundary */
 int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int num_dwords =
 		(ring->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
 	int ret;
@@ -2447,7 +2448,7 @@ int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 	return 0;
 }
 
-void intel_ring_init_seqno(struct intel_engine_cs *engine, u32 seqno)
+void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 
@@ -2533,7 +2534,7 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 			       u32 invalidate, u32 flush)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	uint32_t cmd;
 	int ret;
 
@@ -2579,7 +2580,7 @@ gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	bool ppgtt = USES_PPGTT(req->i915) &&
 			!(dispatch_flags & I915_DISPATCH_SECURE);
 	int ret;
@@ -2605,7 +2606,7 @@ hsw_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			     u64 offset, u32 len,
 			     unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -2630,7 +2631,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 			      u64 offset, u32 len,
 			      unsigned dispatch_flags)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	int ret;
 
 	ret = intel_ring_begin(req, 2);
@@ -2653,7 +2654,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 static int gen6_ring_flush(struct drm_i915_gem_request *req,
 			   u32 invalidate, u32 flush)
 {
-	struct intel_ringbuffer *ring = req->ring;
+	struct intel_ring *ring = req->ring;
 	uint32_t cmd;
 	int ret;
 
@@ -2970,7 +2971,7 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
 }
 
 int
-intel_ring_flush_all_caches(struct drm_i915_gem_request *req)
+intel_engine_flush_all_caches(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
 	int ret;
@@ -2989,7 +2990,7 @@ intel_ring_flush_all_caches(struct drm_i915_gem_request *req)
 }
 
 int
-intel_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
+intel_engine_invalidate_all_caches(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
 	uint32_t flush_domains;
@@ -3009,8 +3010,7 @@ intel_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
 	return 0;
 }
 
-void
-intel_stop_engine(struct intel_engine_cs *engine)
+void intel_engine_stop(struct intel_engine_cs *engine)
 {
 	int ret;
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 1f8e49132c54..5c592d258de0 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -62,7 +62,7 @@ struct  intel_hw_status_page {
 	(i915_gem_obj_ggtt_offset(dev_priv->semaphore_obj) + \
 	 GEN8_SEMAPHORE_OFFSET(from, (__ring)->id))
 
-enum intel_ring_hangcheck_action {
+enum intel_engine_hangcheck_action {
 	HANGCHECK_IDLE = 0,
 	HANGCHECK_WAIT,
 	HANGCHECK_ACTIVE,
@@ -72,17 +72,17 @@ enum intel_ring_hangcheck_action {
 
 #define HANGCHECK_SCORE_RING_HUNG 31
 
-struct intel_ring_hangcheck {
+struct intel_engine_hangcheck {
 	u64 acthd;
 	unsigned long user_interrupts;
 	u32 seqno;
 	int score;
-	enum intel_ring_hangcheck_action action;
+	enum intel_engine_hangcheck_action action;
 	int deadlock;
 	u32 instdone[I915_NUM_INSTDONE_REG];
 };
 
-struct intel_ringbuffer {
+struct intel_ring {
 	struct drm_i915_gem_object *obj;
 	void *vaddr;
 	struct i915_vma *vma;
@@ -149,7 +149,7 @@ struct intel_engine_cs {
 	u64 fence_context;
 	u32		mmio_base;
 	unsigned int irq_shift;
-	struct intel_ringbuffer *buffer;
+	struct intel_ring *buffer;
 	struct list_head buffers;
 
 	/* Rather than have every client wait upon all user interrupts,
@@ -329,7 +329,7 @@ struct intel_engine_cs {
 
 	struct i915_gem_context *last_context;
 
-	struct intel_ring_hangcheck hangcheck;
+	struct intel_engine_hangcheck hangcheck;
 
 	struct {
 		struct drm_i915_gem_object *obj;
@@ -376,8 +376,8 @@ intel_engine_flag(const struct intel_engine_cs *engine)
 }
 
 static inline u32
-intel_ring_sync_index(struct intel_engine_cs *engine,
-		      struct intel_engine_cs *other)
+intel_engine_sync_index(struct intel_engine_cs *engine,
+			struct intel_engine_cs *other)
 {
 	int idx;
 
@@ -439,45 +439,44 @@ intel_write_status_page(struct intel_engine_cs *engine,
 #define I915_GEM_HWS_SCRATCH_INDEX	0x40
 #define I915_GEM_HWS_SCRATCH_ADDR (I915_GEM_HWS_SCRATCH_INDEX << MI_STORE_DWORD_INDEX_SHIFT)
 
-struct intel_ringbuffer *
-intel_engine_create_ringbuffer(struct intel_engine_cs *engine, int size);
-int intel_pin_and_map_ringbuffer_obj(struct drm_i915_private *dev_priv,
-				     struct intel_ringbuffer *ringbuf);
-void intel_unpin_ringbuffer_obj(struct intel_ringbuffer *ringbuf);
-void intel_ringbuffer_free(struct intel_ringbuffer *ring);
+struct intel_ring *
+intel_engine_create_ring(struct intel_engine_cs *engine, int size);
+int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
+			   struct intel_ring *ring);
+void intel_unpin_ring(struct intel_ring *ring);
+void intel_ring_free(struct intel_ring *ring);
 
-void intel_stop_engine(struct intel_engine_cs *engine);
-void intel_cleanup_engine(struct intel_engine_cs *engine);
+void intel_engine_stop(struct intel_engine_cs *engine);
+void intel_engine_cleanup(struct intel_engine_cs *engine);
 
 int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request);
 
 int __must_check intel_ring_begin(struct drm_i915_gem_request *req, int n);
 int __must_check intel_ring_cacheline_align(struct drm_i915_gem_request *req);
 
-static inline void intel_ring_emit(struct intel_ringbuffer *ring, u32 data)
+static inline void intel_ring_emit(struct intel_ring *ring, u32 data)
 {
 	*(uint32_t *)(ring->vaddr + ring->tail) = data;
 	ring->tail += 4;
 }
 
-static inline void intel_ring_emit_reg(struct intel_ringbuffer *ring,
-				       i915_reg_t reg)
+static inline void intel_ring_emit_reg(struct intel_ring *ring, i915_reg_t reg)
 {
 	intel_ring_emit(ring, i915_mmio_reg_offset(reg));
 }
 
-static inline void intel_ring_advance(struct intel_ringbuffer *ring)
+static inline void intel_ring_advance(struct intel_ring *ring)
 {
 	ring->tail &= ring->size - 1;
 }
 
 int __intel_ring_space(int head, int tail, int size);
-void intel_ring_update_space(struct intel_ringbuffer *ringbuf);
+void intel_ring_update_space(struct intel_ring *ringbuf);
 
 int __must_check intel_engine_idle(struct intel_engine_cs *engine);
-void intel_ring_init_seqno(struct intel_engine_cs *engine, u32 seqno);
-int intel_ring_flush_all_caches(struct drm_i915_gem_request *req);
-int intel_ring_invalidate_all_caches(struct drm_i915_gem_request *req);
+void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno);
+int intel_engine_flush_all_caches(struct drm_i915_gem_request *req);
+int intel_engine_invalidate_all_caches(struct drm_i915_gem_request *req);
 
 int intel_init_pipe_control(struct intel_engine_cs *engine, int size);
 void intel_fini_pipe_control(struct intel_engine_cs *engine);
@@ -491,7 +490,7 @@ int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine);
 
-u64 intel_ring_get_active_head(struct intel_engine_cs *engine);
+u64 intel_engine_get_active_head(struct intel_engine_cs *engine);
 static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine)
 {
 	return intel_read_status_page(engine, I915_GEM_HWS_INDEX);
@@ -499,7 +498,7 @@ static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine)
 
 int init_workarounds_ring(struct intel_engine_cs *engine);
 
-static inline u32 intel_ring_get_tail(struct intel_ringbuffer *ringbuf)
+static inline u32 intel_ring_get_tail(struct intel_ring *ringbuf)
 {
 	return ringbuf->tail;
 }
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 15/31] drm/i915: Rename residual ringbuf parameters

[Chris Wilson]

Now that we have a clear ring/engine split and a struct intel_ring, we
no longer need the stopgap ringbuf names.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 66 ++++++++++++++++-----------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  6 +--
 2 files changed, 36 insertions(+), 36 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 90572b862eec..5ff44eb46f55 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -47,15 +47,15 @@ int __intel_ring_space(int head, int tail, int size)
 	return space - I915_RING_FREE_SPACE;
 }
 
-void intel_ring_update_space(struct intel_ring *ringbuf)
+void intel_ring_update_space(struct intel_ring *ring)
 {
-	if (ringbuf->last_retired_head != -1) {
-		ringbuf->head = ringbuf->last_retired_head;
-		ringbuf->last_retired_head = -1;
+	if (ring->last_retired_head != -1) {
+		ring->head = ring->last_retired_head;
+		ring->last_retired_head = -1;
 	}
 
-	ringbuf->space = __intel_ring_space(ringbuf->head & HEAD_ADDR,
-					    ringbuf->tail, ringbuf->size);
+	ring->space = __intel_ring_space(ring->head & HEAD_ADDR,
+					 ring->tail, ring->size);
 }
 
 static void __intel_engine_submit(struct intel_engine_cs *engine)
@@ -1993,25 +1993,25 @@ static int init_phys_status_page(struct intel_engine_cs *engine)
 	return 0;
 }
 
-void intel_unpin_ring(struct intel_ring *ringbuf)
+void intel_unpin_ring(struct intel_ring *ring)
 {
-	GEM_BUG_ON(!ringbuf->vma);
-	GEM_BUG_ON(!ringbuf->vaddr);
+	GEM_BUG_ON(!ring->vma);
+	GEM_BUG_ON(!ring->vaddr);
 
-	if (HAS_LLC(ringbuf->obj->base.dev) && !ringbuf->obj->stolen)
-		i915_gem_object_unpin_map(ringbuf->obj);
+	if (HAS_LLC(ring->obj->base.dev) && !ring->obj->stolen)
+		i915_gem_object_unpin_map(ring->obj);
 	else
-		i915_vma_unpin_iomap(ringbuf->vma);
-	ringbuf->vaddr = NULL;
+		i915_vma_unpin_iomap(ring->vma);
+	ring->vaddr = NULL;
 
-	i915_gem_object_ggtt_unpin(ringbuf->obj);
-	ringbuf->vma = NULL;
+	i915_gem_object_ggtt_unpin(ring->obj);
+	ring->vma = NULL;
 }
 
 int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
-			   struct intel_ring *ringbuf)
+			   struct intel_ring *ring)
 {
-	struct drm_i915_gem_object *obj = ringbuf->obj;
+	struct drm_i915_gem_object *obj = ring->obj;
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
 	unsigned flags = PIN_OFFSET_BIAS | 4096;
 	void *addr;
@@ -2052,8 +2052,8 @@ int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
 		}
 	}
 
-	ringbuf->vaddr = addr;
-	ringbuf->vma = i915_gem_obj_to_ggtt(obj);
+	ring->vaddr = addr;
+	ring->vma = i915_gem_obj_to_ggtt(obj);
 	return 0;
 
 err_unpin:
@@ -2061,29 +2061,29 @@ err_unpin:
 	return ret;
 }
 
-static void intel_destroy_ringbuffer_obj(struct intel_ring *ringbuf)
+static void intel_destroy_ringbuffer_obj(struct intel_ring *ring)
 {
-	i915_gem_object_put(ringbuf->obj);
-	ringbuf->obj = NULL;
+	i915_gem_object_put(ring->obj);
+	ring->obj = NULL;
 }
 
 static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
-				      struct intel_ring *ringbuf)
+				      struct intel_ring *ring)
 {
 	struct drm_i915_gem_object *obj;
 
 	obj = NULL;
 	if (!HAS_LLC(dev))
-		obj = i915_gem_object_create_stolen(dev, ringbuf->size);
+		obj = i915_gem_object_create_stolen(dev, ring->size);
 	if (obj == NULL)
-		obj = i915_gem_object_create(dev, ringbuf->size);
+		obj = i915_gem_object_create(dev, ring->size);
 	if (IS_ERR(obj))
 		return PTR_ERR(obj);
 
 	/* mark ring buffers as read-only from GPU side by default */
 	obj->gt_ro = 1;
 
-	ringbuf->obj = obj;
+	ring->obj = obj;
 
 	return 0;
 }
@@ -2190,7 +2190,7 @@ static void intel_ring_context_unpin(struct i915_gem_context *ctx,
 static int intel_init_engine(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
-	struct intel_ring *ringbuf;
+	struct intel_ring *ring;
 	int ret;
 
 	WARN_ON(engine->buffer);
@@ -2215,12 +2215,12 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 	if (ret)
 		goto error;
 
-	ringbuf = intel_engine_create_ring(engine, 32 * PAGE_SIZE);
-	if (IS_ERR(ringbuf)) {
-		ret = PTR_ERR(ringbuf);
+	ring = intel_engine_create_ring(engine, 32 * PAGE_SIZE);
+	if (IS_ERR(ring)) {
+		ret = PTR_ERR(ring);
 		goto error;
 	}
-	engine->buffer = ringbuf;
+	engine->buffer = ring;
 
 	if (I915_NEED_GFX_HWS(dev_priv)) {
 		ret = init_status_page(engine);
@@ -2233,11 +2233,11 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 			goto error;
 	}
 
-	ret = intel_pin_and_map_ring(dev_priv, ringbuf);
+	ret = intel_pin_and_map_ring(dev_priv, ring);
 	if (ret) {
 		DRM_ERROR("Failed to pin and map ringbuffer %s: %d\n",
 				engine->name, ret);
-		intel_destroy_ringbuffer_obj(ringbuf);
+		intel_destroy_ringbuffer_obj(ring);
 		goto error;
 	}
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 5c592d258de0..a46b4de0847c 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -471,7 +471,7 @@ static inline void intel_ring_advance(struct intel_ring *ring)
 }
 
 int __intel_ring_space(int head, int tail, int size);
-void intel_ring_update_space(struct intel_ring *ringbuf);
+void intel_ring_update_space(struct intel_ring *ring);
 
 int __must_check intel_engine_idle(struct intel_engine_cs *engine);
 void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno);
@@ -498,9 +498,9 @@ static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine)
 
 int init_workarounds_ring(struct intel_engine_cs *engine);
 
-static inline u32 intel_ring_get_tail(struct intel_ring *ringbuf)
+static inline u32 intel_ring_get_tail(struct intel_ring *ring)
 {
-	return ringbuf->tail;
+	return ring->tail;
 }
 
 /*
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 16/31] drm/i915: Rename intel_pin_and_map_ring()

[Chris Wilson]

For more consistent oop-naming, we would use intel_ring_verb, so pick
intel_ring_pin() and intel_ring_unpin().

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_lrc.c        |  4 ++--
 drivers/gpu/drm/i915/intel_ringbuffer.c | 38 ++++++++++++++++-----------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  5 ++---
 3 files changed, 23 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index beb82ef1c321..08838c0286b9 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -973,7 +973,7 @@ static int intel_lr_context_pin(struct i915_gem_context *ctx,
 
 	lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
 
-	ret = intel_pin_and_map_ring(dev_priv, ce->ring);
+	ret = intel_ring_pin(ce->ring);
 	if (ret)
 		goto unpin_map;
 
@@ -1011,7 +1011,7 @@ void intel_lr_context_unpin(struct i915_gem_context *ctx,
 	if (--ce->pin_count)
 		return;
 
-	intel_unpin_ring(ce->ring);
+	intel_ring_unpin(ce->ring);
 
 	i915_gem_object_unpin_map(ce->state);
 	i915_gem_object_ggtt_unpin(ce->state);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 5ff44eb46f55..b7253dcf9279 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1993,24 +1993,9 @@ static int init_phys_status_page(struct intel_engine_cs *engine)
 	return 0;
 }
 
-void intel_unpin_ring(struct intel_ring *ring)
-{
-	GEM_BUG_ON(!ring->vma);
-	GEM_BUG_ON(!ring->vaddr);
-
-	if (HAS_LLC(ring->obj->base.dev) && !ring->obj->stolen)
-		i915_gem_object_unpin_map(ring->obj);
-	else
-		i915_vma_unpin_iomap(ring->vma);
-	ring->vaddr = NULL;
-
-	i915_gem_object_ggtt_unpin(ring->obj);
-	ring->vma = NULL;
-}
-
-int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
-			   struct intel_ring *ring)
+int intel_ring_pin(struct intel_ring *ring)
 {
+	struct drm_i915_private *dev_priv = ring->engine->i915;
 	struct drm_i915_gem_object *obj = ring->obj;
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
 	unsigned flags = PIN_OFFSET_BIAS | 4096;
@@ -2061,6 +2046,21 @@ err_unpin:
 	return ret;
 }
 
+void intel_ring_unpin(struct intel_ring *ring)
+{
+	GEM_BUG_ON(!ring->vma);
+	GEM_BUG_ON(!ring->vaddr);
+
+	if (HAS_LLC(ring->engine->i915) && !ring->obj->stolen)
+		i915_gem_object_unpin_map(ring->obj);
+	else
+		i915_vma_unpin_iomap(ring->vma);
+	ring->vaddr = NULL;
+
+	i915_gem_object_ggtt_unpin(ring->obj);
+	ring->vma = NULL;
+}
+
 static void intel_destroy_ringbuffer_obj(struct intel_ring *ring)
 {
 	i915_gem_object_put(ring->obj);
@@ -2233,7 +2233,7 @@ static int intel_init_engine(struct intel_engine_cs *engine)
 			goto error;
 	}
 
-	ret = intel_pin_and_map_ring(dev_priv, ring);
+	ret = intel_ring_pin(ring);
 	if (ret) {
 		DRM_ERROR("Failed to pin and map ringbuffer %s: %d\n",
 				engine->name, ret);
@@ -2261,7 +2261,7 @@ void intel_engine_cleanup(struct intel_engine_cs *engine)
 		intel_engine_stop(engine);
 		WARN_ON(!IS_GEN2(dev_priv) && (I915_READ_MODE(engine) & MODE_IDLE) == 0);
 
-		intel_unpin_ring(engine->buffer);
+		intel_ring_unpin(engine->buffer);
 		intel_ring_free(engine->buffer);
 		engine->buffer = NULL;
 	}
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index a46b4de0847c..6c74d52812c0 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -441,9 +441,8 @@ intel_write_status_page(struct intel_engine_cs *engine,
 
 struct intel_ring *
 intel_engine_create_ring(struct intel_engine_cs *engine, int size);
-int intel_pin_and_map_ring(struct drm_i915_private *dev_priv,
-			   struct intel_ring *ring);
-void intel_unpin_ring(struct intel_ring *ring);
+int intel_ring_pin(struct intel_ring *ring);
+void intel_ring_unpin(struct intel_ring *ring);
 void intel_ring_free(struct intel_ring *ring);
 
 void intel_engine_stop(struct intel_engine_cs *engine);
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Chris Wilson]

Space for flushing the GPU cache prior to completing the request is
preallocated and so cannot fail.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  9 +---
 drivers/gpu/drm/i915/i915_gem_gtt.c        | 11 +++--
 drivers/gpu/drm/i915/i915_gem_request.c    |  7 ++-
 drivers/gpu/drm/i915/intel_lrc.c           | 47 +++----------------
 drivers/gpu/drm/i915/intel_lrc.h           |  2 -
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 72 +++++++-----------------------
 drivers/gpu/drm/i915/intel_ringbuffer.h    |  7 ---
 8 files changed, 37 insertions(+), 120 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 3336a5fcd029..beece8feb8fe 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -568,7 +568,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 	 * itlb_before_ctx_switch.
 	 */
 	if (IS_GEN6(dev_priv)) {
-		ret = engine->flush(req, I915_GEM_GPU_DOMAINS, 0);
+		ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index ca1d4f573832..048050176ff9 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -998,10 +998,8 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 	if (flush_domains & I915_GEM_DOMAIN_GTT)
 		wmb();
 
-	/* Unconditionally invalidate gpu caches and ensure that we do flush
-	 * any residual writes from the previous batch.
-	 */
-	return intel_engine_invalidate_all_caches(req);
+	/* Unconditionally invalidate gpu caches and TLBs. */
+	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
 }
 
 static bool
@@ -1163,9 +1161,6 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 static void
 i915_gem_execbuffer_retire_commands(struct i915_execbuffer_params *params)
 {
-	/* Unconditionally force add_request to emit a full flush. */
-	params->engine->gpu_caches_dirty = true;
-
 	/* Add a breadcrumb for the completion of the batch buffer */
 	__i915_add_request(params->request, params->batch_obj, true);
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index ebfa0406a6a1..39fa9eb10514 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1666,7 +1666,8 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	int ret;
 
 	/* NB: TLBs must be flushed and invalidated before a switch */
-	ret = engine->flush(req, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	ret = engine->emit_flush(req,
+				 I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
@@ -1693,7 +1694,8 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	int ret;
 
 	/* NB: TLBs must be flushed and invalidated before a switch */
-	ret = engine->flush(req, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	ret = engine->emit_flush(req,
+				 I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
@@ -1711,8 +1713,9 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 
 	/* XXX: RCS is the only one to auto invalidate the TLBs? */
 	if (engine->id != RCS) {
-		ret = engine->flush(req,
-				    I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+		ret = engine->emit_flush(req,
+					 I915_GEM_GPU_DOMAINS,
+					 I915_GEM_GPU_DOMAINS);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 942b5b1f1602..7b772d914e23 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -451,10 +451,9 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 * what.
 	 */
 	if (flush_caches) {
-		if (i915.enable_execlists)
-			ret = logical_ring_flush_all_caches(request);
-		else
-			ret = intel_engine_flush_all_caches(request);
+		ret = request->engine->emit_flush(request,
+						  0, I915_GEM_GPU_DOMAINS);
+
 		/* Not allowed to fail! */
 		WARN(ret, "*_ring_flush_all_caches failed: %d!\n", ret);
 	}
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 08838c0286b9..6487794ac96c 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -642,24 +642,6 @@ static void execlists_context_queue(struct drm_i915_gem_request *request)
 	spin_unlock_bh(&engine->execlist_lock);
 }
 
-static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
-{
-	struct intel_engine_cs *engine = req->engine;
-	uint32_t flush_domains;
-	int ret;
-
-	flush_domains = 0;
-	if (engine->gpu_caches_dirty)
-		flush_domains = I915_GEM_GPU_DOMAINS;
-
-	ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
-	if (ret)
-		return ret;
-
-	engine->gpu_caches_dirty = false;
-	return 0;
-}
-
 static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
 				 struct list_head *vmas)
 {
@@ -690,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
 	/* Unconditionally invalidate gpu caches and ensure that we do flush
 	 * any residual writes from the previous batch.
 	 */
-	return logical_ring_invalidate_all_caches(req);
+	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
 }
 
 int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
@@ -930,22 +912,6 @@ void intel_logical_ring_stop(struct intel_engine_cs *engine)
 	I915_WRITE_MODE(engine, _MASKED_BIT_DISABLE(STOP_RING));
 }
 
-int logical_ring_flush_all_caches(struct drm_i915_gem_request *req)
-{
-	struct intel_engine_cs *engine = req->engine;
-	int ret;
-
-	if (!engine->gpu_caches_dirty)
-		return 0;
-
-	ret = engine->emit_flush(req, 0, I915_GEM_GPU_DOMAINS);
-	if (ret)
-		return ret;
-
-	engine->gpu_caches_dirty = false;
-	return 0;
-}
-
 static int intel_lr_context_pin(struct i915_gem_context *ctx,
 				struct intel_engine_cs *engine)
 {
@@ -1026,15 +992,15 @@ void intel_lr_context_unpin(struct i915_gem_context *ctx,
 static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 {
 	int ret, i;
-	struct intel_engine_cs *engine = req->engine;
 	struct intel_ring *ring = req->ring;
 	struct i915_workarounds *w = &req->i915->workarounds;
 
 	if (w->count == 0)
 		return 0;
 
-	engine->gpu_caches_dirty = true;
-	ret = logical_ring_flush_all_caches(req);
+	ret = req->engine->emit_flush(req,
+				      I915_GEM_GPU_DOMAINS,
+				      I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
@@ -1051,8 +1017,9 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 
 	intel_ring_advance(ring);
 
-	engine->gpu_caches_dirty = true;
-	ret = logical_ring_flush_all_caches(req);
+	ret = req->engine->emit_flush(req,
+				      I915_GEM_GPU_DOMAINS,
+				      I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index d26fb44549e5..33e0193e5451 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -72,8 +72,6 @@ int logical_xcs_ring_init(struct intel_engine_cs *engine);
 
 int intel_engines_init(struct drm_device *dev);
 
-int logical_ring_flush_all_caches(struct drm_i915_gem_request *req);
-
 /* Logical Ring Contexts */
 
 /* One extra page is added before LRC for GuC as shared data */
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index b7253dcf9279..bd3ed9e12afe 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -688,8 +688,9 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (w->count == 0)
 		return 0;
 
-	req->engine->gpu_caches_dirty = true;
-	ret = intel_engine_flush_all_caches(req);
+	ret = req->engine->emit_flush(req,
+				      I915_GEM_GPU_DOMAINS,
+				      I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
@@ -706,8 +707,9 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 
 	intel_ring_advance(ring);
 
-	req->engine->gpu_caches_dirty = true;
-	ret = intel_engine_flush_all_caches(req);
+	ret = req->engine->emit_flush(req,
+				      I915_GEM_GPU_DOMAINS,
+				      I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
 
@@ -2860,21 +2862,21 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
 	if (INTEL_GEN(dev_priv) >= 8) {
 		engine->init_context = intel_rcs_ctx_init;
 		engine->add_request = gen8_render_add_request;
-		engine->flush = gen8_render_ring_flush;
+		engine->emit_flush = gen8_render_ring_flush;
 		if (i915.semaphores)
 			engine->semaphore.signal = gen8_rcs_signal;
 	} else if (INTEL_GEN(dev_priv) >= 6) {
 		engine->init_context = intel_rcs_ctx_init;
-		engine->flush = gen7_render_ring_flush;
+		engine->emit_flush = gen7_render_ring_flush;
 		if (IS_GEN6(dev_priv))
-			engine->flush = gen6_render_ring_flush;
+			engine->emit_flush = gen6_render_ring_flush;
 	} else if (IS_GEN5(dev_priv)) {
-		engine->flush = gen4_render_ring_flush;
+		engine->emit_flush = gen4_render_ring_flush;
 	} else {
 		if (INTEL_GEN(dev_priv) < 4)
-			engine->flush = gen2_render_ring_flush;
+			engine->emit_flush = gen2_render_ring_flush;
 		else
-			engine->flush = gen4_render_ring_flush;
+			engine->emit_flush = gen4_render_ring_flush;
 		engine->irq_enable_mask = I915_USER_INTERRUPT;
 	}
 
@@ -2911,12 +2913,12 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 		/* gen6 bsd needs a special wa for tail updates */
 		if (IS_GEN6(dev_priv))
 			engine->write_tail = gen6_bsd_ring_write_tail;
-		engine->flush = gen6_bsd_ring_flush;
+		engine->emit_flush = gen6_bsd_ring_flush;
 		if (INTEL_GEN(dev_priv) < 8)
 			engine->irq_enable_mask = GT_BSD_USER_INTERRUPT;
 	} else {
 		engine->mmio_base = BSD_RING_BASE;
-		engine->flush = bsd_ring_flush;
+		engine->emit_flush = bsd_ring_flush;
 		if (IS_GEN5(dev_priv))
 			engine->irq_enable_mask = ILK_BSD_USER_INTERRUPT;
 		else
@@ -2935,7 +2937,7 @@ int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine)
 
 	intel_ring_default_vfuncs(dev_priv, engine);
 
-	engine->flush = gen6_bsd_ring_flush;
+	engine->emit_flush = gen6_bsd_ring_flush;
 
 	return intel_init_engine(engine);
 }
@@ -2946,7 +2948,7 @@ int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
 
 	intel_ring_default_vfuncs(dev_priv, engine);
 
-	engine->flush = gen6_ring_flush;
+	engine->emit_flush = gen6_ring_flush;
 	if (INTEL_GEN(dev_priv) < 8)
 		engine->irq_enable_mask = GT_BLT_USER_INTERRUPT;
 
@@ -2959,7 +2961,7 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
 
 	intel_ring_default_vfuncs(dev_priv, engine);
 
-	engine->flush = gen6_ring_flush;
+	engine->emit_flush = gen6_ring_flush;
 
 	if (INTEL_GEN(dev_priv) < 8) {
 		engine->irq_enable_mask = PM_VEBOX_USER_INTERRUPT;
@@ -2970,46 +2972,6 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
 	return intel_init_engine(engine);
 }
 
-int
-intel_engine_flush_all_caches(struct drm_i915_gem_request *req)
-{
-	struct intel_engine_cs *engine = req->engine;
-	int ret;
-
-	if (!engine->gpu_caches_dirty)
-		return 0;
-
-	ret = engine->flush(req, 0, I915_GEM_GPU_DOMAINS);
-	if (ret)
-		return ret;
-
-	trace_i915_gem_ring_flush(req, 0, I915_GEM_GPU_DOMAINS);
-
-	engine->gpu_caches_dirty = false;
-	return 0;
-}
-
-int
-intel_engine_invalidate_all_caches(struct drm_i915_gem_request *req)
-{
-	struct intel_engine_cs *engine = req->engine;
-	uint32_t flush_domains;
-	int ret;
-
-	flush_domains = 0;
-	if (engine->gpu_caches_dirty)
-		flush_domains = I915_GEM_GPU_DOMAINS;
-
-	ret = engine->flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
-	if (ret)
-		return ret;
-
-	trace_i915_gem_ring_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
-
-	engine->gpu_caches_dirty = false;
-	return 0;
-}
-
 void intel_engine_stop(struct intel_engine_cs *engine)
 {
 	int ret;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 6c74d52812c0..05ac78a7ea64 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -206,9 +206,6 @@ struct intel_engine_cs {
 
 	void		(*write_tail)(struct intel_engine_cs *engine,
 				      u32 value);
-	int __must_check (*flush)(struct drm_i915_gem_request *req,
-				  u32	invalidate_domains,
-				  u32	flush_domains);
 	int		(*add_request)(struct drm_i915_gem_request *req);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
@@ -325,8 +322,6 @@ struct intel_engine_cs {
 	 */
 	u32 last_submitted_seqno;
 
-	bool gpu_caches_dirty;
-
 	struct i915_gem_context *last_context;
 
 	struct intel_engine_hangcheck hangcheck;
@@ -474,8 +469,6 @@ void intel_ring_update_space(struct intel_ring *ring);
 
 int __must_check intel_engine_idle(struct intel_engine_cs *engine);
 void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno);
-int intel_engine_flush_all_caches(struct drm_i915_gem_request *req);
-int intel_engine_invalidate_all_caches(struct drm_i915_gem_request *req);
 
 int intel_init_pipe_control(struct intel_engine_cs *engine, int size);
 void intel_fini_pipe_control(struct intel_engine_cs *engine);
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Chris Wilson]

If is simpler and leads to more readable code through the callstack if
the allocation returns the allocated struct through the return value.

The importance of this is that it no longer looks like we accidentally
allocate requests as side-effect of calling certain functions.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_drv.h            |  3 +-
 drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
 drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
 drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
 drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
 drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
 drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
 8 files changed, 79 insertions(+), 140 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 09e72b0510ab..1031f5a7126a 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -3171,8 +3171,7 @@ static inline void i915_gem_object_unpin_map(struct drm_i915_gem_object *obj)
 
 int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
 int i915_gem_object_sync(struct drm_i915_gem_object *obj,
-			 struct intel_engine_cs *to,
-			 struct drm_i915_gem_request **to_req);
+			 struct drm_i915_gem_request *to);
 void i915_vma_move_to_active(struct i915_vma *vma,
 			     struct drm_i915_gem_request *req);
 int i915_gem_dumb_create(struct drm_file *file_priv,
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 59890f523c5f..b6c4ff63725f 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2845,51 +2845,35 @@ out:
 
 static int
 __i915_gem_object_sync(struct drm_i915_gem_object *obj,
-		       struct intel_engine_cs *to,
-		       struct drm_i915_gem_request *from_req,
-		       struct drm_i915_gem_request **to_req)
+		       struct drm_i915_gem_request *to,
+		       struct drm_i915_gem_request *from)
 {
-	struct intel_engine_cs *from;
 	int ret;
 
-	from = i915_gem_request_get_engine(from_req);
-	if (to == from)
+	if (to->engine == from->engine)
 		return 0;
 
-	if (i915_gem_request_completed(from_req))
+	if (i915_gem_request_completed(from))
 		return 0;
 
 	if (!i915.semaphores) {
-		struct drm_i915_private *i915 = to_i915(obj->base.dev);
-		ret = __i915_wait_request(from_req,
-					  i915->mm.interruptible,
+		ret = __i915_wait_request(from,
+					  from->i915->mm.interruptible,
 					  NULL,
 					  NO_WAITBOOST);
 		if (ret)
 			return ret;
 
-		i915_gem_object_retire_request(obj, from_req);
+		i915_gem_object_retire_request(obj, from);
 	} else {
-		int idx = intel_engine_sync_index(from, to);
-		u32 seqno = i915_gem_request_get_seqno(from_req);
+		int idx = intel_engine_sync_index(from->engine, to->engine);
+		u32 seqno = i915_gem_request_get_seqno(from);
 
-		WARN_ON(!to_req);
-
-		if (seqno <= from->semaphore.sync_seqno[idx])
+		if (seqno <= from->engine->semaphore.sync_seqno[idx])
 			return 0;
 
-		if (*to_req == NULL) {
-			struct drm_i915_gem_request *req;
-
-			req = i915_gem_request_alloc(to, NULL);
-			if (IS_ERR(req))
-				return PTR_ERR(req);
-
-			*to_req = req;
-		}
-
-		trace_i915_gem_ring_sync_to(*to_req, from, from_req);
-		ret = to->semaphore.sync_to(*to_req, from, seqno);
+		trace_i915_gem_ring_sync_to(to, from);
+		ret = to->engine->semaphore.sync_to(to, from->engine, seqno);
 		if (ret)
 			return ret;
 
@@ -2897,8 +2881,8 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
 		 * might have just caused seqno wrap under
 		 * the radar.
 		 */
-		from->semaphore.sync_seqno[idx] =
-			i915_gem_request_get_seqno(obj->last_read_req[from->id]);
+		from->engine->semaphore.sync_seqno[idx] =
+			i915_gem_request_get_seqno(obj->last_read_req[from->engine->id]);
 	}
 
 	return 0;
@@ -2908,17 +2892,12 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
  * i915_gem_object_sync - sync an object to a ring.
  *
  * @obj: object which may be in use on another ring.
- * @to: ring we wish to use the object on. May be NULL.
- * @to_req: request we wish to use the object for. See below.
- *          This will be allocated and returned if a request is
- *          required but not passed in.
+ * @to: request we are wishing to use
  *
  * This code is meant to abstract object synchronization with the GPU.
- * Calling with NULL implies synchronizing the object with the CPU
- * rather than a particular GPU ring. Conceptually we serialise writes
- * between engines inside the GPU. We only allow one engine to write
- * into a buffer at any time, but multiple readers. To ensure each has
- * a coherent view of memory, we must:
+ * Conceptually we serialise writes between engines inside the GPU.
+ * We only allow one engine to write into a buffer at any time, but
+ * multiple readers. To ensure each has a coherent view of memory, we must:
  *
  * - If there is an outstanding write request to the object, the new
  *   request must wait for it to complete (either CPU or in hw, requests
@@ -2927,22 +2906,11 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
  * - If we are a write request (pending_write_domain is set), the new
  *   request must wait for outstanding read requests to complete.
  *
- * For CPU synchronisation (NULL to) no request is required. For syncing with
- * rings to_req must be non-NULL. However, a request does not have to be
- * pre-allocated. If *to_req is NULL and sync commands will be emitted then a
- * request will be allocated automatically and returned through *to_req. Note
- * that it is not guaranteed that commands will be emitted (because the system
- * might already be idle). Hence there is no need to create a request that
- * might never have any work submitted. Note further that if a request is
- * returned in *to_req, it is the responsibility of the caller to submit
- * that request (after potentially adding more work to it).
- *
  * Returns 0 if successful, else propagates up the lower layer error.
  */
 int
 i915_gem_object_sync(struct drm_i915_gem_object *obj,
-		     struct intel_engine_cs *to,
-		     struct drm_i915_gem_request **to_req)
+		     struct drm_i915_gem_request *to)
 {
 	const bool readonly = obj->base.pending_write_domain == 0;
 	struct drm_i915_gem_request *req[I915_NUM_ENGINES];
@@ -2951,9 +2919,6 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 	if (!obj->active)
 		return 0;
 
-	if (to == NULL)
-		return i915_gem_object_wait_rendering(obj, readonly);
-
 	n = 0;
 	if (readonly) {
 		if (obj->last_write_req)
@@ -2964,7 +2929,7 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 				req[n++] = obj->last_read_req[i];
 	}
 	for (i = 0; i < n; i++) {
-		ret = __i915_gem_object_sync(obj, to, req[i], to_req);
+		ret = __i915_gem_object_sync(obj, to, req[i]);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 048050176ff9..75d19932847d 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -981,7 +981,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 		struct drm_i915_gem_object *obj = vma->obj;
 
 		if (obj->active & other_rings) {
-			ret = i915_gem_object_sync(obj, req->engine, &req);
+			ret = i915_gem_object_sync(obj, req);
 			if (ret)
 				return ret;
 		}
@@ -1427,7 +1427,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct i915_ggtt *ggtt = &dev_priv->ggtt;
-	struct drm_i915_gem_request *req = NULL;
 	struct eb_vmas *eb;
 	struct drm_i915_gem_object *batch_obj;
 	struct drm_i915_gem_exec_object2 shadow_exec_entry;
@@ -1615,13 +1614,13 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		params->batch_obj_vm_offset = i915_gem_obj_offset(batch_obj, vm);
 
 	/* Allocate a request for this batch buffer nice and early. */
-	req = i915_gem_request_alloc(engine, ctx);
-	if (IS_ERR(req)) {
-		ret = PTR_ERR(req);
+	params->request = i915_gem_request_alloc(engine, ctx);
+	if (IS_ERR(params->request)) {
+		ret = PTR_ERR(params->request);
 		goto err_batch_unpin;
 	}
 
-	ret = i915_gem_request_add_to_client(req, file);
+	ret = i915_gem_request_add_to_client(params->request, file);
 	if (ret)
 		goto err_request;
 
@@ -1637,7 +1636,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	params->dispatch_flags          = dispatch_flags;
 	params->batch_obj               = batch_obj;
 	params->ctx                     = ctx;
-	params->request                 = req;
 
 	ret = dev_priv->gt.execbuf_submit(params, args, &eb->vmas);
 err_request:
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 7b772d914e23..995ae99e9cf7 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -292,10 +292,21 @@ static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno)
 	return 0;
 }
 
-static inline int
-__i915_gem_request_alloc(struct intel_engine_cs *engine,
-			 struct i915_gem_context *ctx,
-			 struct drm_i915_gem_request **req_out)
+/**
+ * i915_gem_request_alloc - allocate a request structure
+ *
+ * @engine: engine that we wish to issue the request on.
+ * @ctx: context that the request will be associated with.
+ *       This can be NULL if the request is not directly related to
+ *       any specific user context, in which case this function will
+ *       choose an appropriate context to use.
+ *
+ * Returns a pointer to the allocated request if successful,
+ * or an error code if not.
+ */
+struct drm_i915_gem_request *
+i915_gem_request_alloc(struct intel_engine_cs *engine,
+		       struct i915_gem_context *ctx)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	unsigned int reset_counter = i915_reset_counter(&dev_priv->gpu_error);
@@ -303,18 +314,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
 	u32 seqno;
 	int ret;
 
-	if (!req_out)
-		return -EINVAL;
-
-	*req_out = NULL;
-
 	/* ABI: Before userspace accesses the GPU (e.g. execbuffer), report
 	 * EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex
 	 * and restart.
 	 */
 	ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible);
 	if (ret)
-		return ret;
+		return ERR_PTR(ret);
 
 	/* Move the oldest request to the slab-cache (if not in use!) */
 	req = list_first_entry_or_null(&engine->request_list,
@@ -324,7 +330,7 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
 
 	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
 	if (!req)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
 	ret = i915_gem_get_seqno(dev_priv, &seqno);
 	if (ret)
@@ -357,39 +363,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
 	if (ret)
 		goto err_ctx;
 
-	*req_out = req;
-	return 0;
+	return req;
 
 err_ctx:
 	i915_gem_context_put(ctx);
 err:
 	kmem_cache_free(dev_priv->requests, req);
-	return ret;
-}
-
-/**
- * i915_gem_request_alloc - allocate a request structure
- *
- * @engine: engine that we wish to issue the request on.
- * @ctx: context that the request will be associated with.
- *       This can be NULL if the request is not directly related to
- *       any specific user context, in which case this function will
- *       choose an appropriate context to use.
- *
- * Returns a pointer to the allocated request if successful,
- * or an error code if not.
- */
-struct drm_i915_gem_request *
-i915_gem_request_alloc(struct intel_engine_cs *engine,
-		       struct i915_gem_context *ctx)
-{
-	struct drm_i915_gem_request *req;
-	int err;
-
-	if (!ctx)
-		ctx = engine->i915->kernel_context;
-	err = __i915_gem_request_alloc(engine, ctx, &req);
-	return err ? ERR_PTR(err) : req;
+	return ERR_PTR(ret);
 }
 
 static void i915_gem_mark_busy(const struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 007112d1e049..9e43c0aa6e3b 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -449,10 +449,9 @@ TRACE_EVENT(i915_gem_evict_vm,
 );
 
 TRACE_EVENT(i915_gem_ring_sync_to,
-	    TP_PROTO(struct drm_i915_gem_request *to_req,
-		     struct intel_engine_cs *from,
-		     struct drm_i915_gem_request *req),
-	    TP_ARGS(to_req, from, req),
+	    TP_PROTO(struct drm_i915_gem_request *to,
+		     struct drm_i915_gem_request *from),
+	    TP_ARGS(to, from),
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
@@ -463,9 +462,9 @@ TRACE_EVENT(i915_gem_ring_sync_to,
 
 	    TP_fast_assign(
 			   __entry->dev = from->i915->drm.primary->index;
-			   __entry->sync_from = from->id;
-			   __entry->sync_to = to_req->engine->id;
-			   __entry->seqno = req->fence.seqno;
+			   __entry->sync_from = from->engine->id;
+			   __entry->sync_to = to->engine->id;
+			   __entry->seqno = from->fence.seqno;
 			   ),
 
 	    TP_printk("dev=%u, sync-from=%u, sync-to=%u, seqno=%u",
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index bff172c45ff7..5d4420b67642 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -11583,7 +11583,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	struct intel_flip_work *work;
 	struct intel_engine_cs *engine;
 	bool mmio_flip;
-	struct drm_i915_gem_request *request = NULL;
+	struct drm_i915_gem_request *request;
 	int ret;
 
 	/*
@@ -11690,22 +11690,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 	mmio_flip = use_mmio_flip(engine, obj);
 
-	/* When using CS flips, we want to emit semaphores between rings.
-	 * However, when using mmio flips we will create a task to do the
-	 * synchronisation, so all we want here is to pin the framebuffer
-	 * into the display plane and skip any waits.
-	 */
-	if (!mmio_flip) {
-		ret = i915_gem_object_sync(obj, engine, &request);
-		if (!ret && !request) {
-			request = i915_gem_request_alloc(engine, NULL);
-			ret = PTR_ERR_OR_ZERO(request);
-		}
-
-		if (ret)
-			goto cleanup_pending;
-	}
-
 	ret = intel_pin_and_fence_fb_obj(fb, primary->state->rotation);
 	if (ret)
 		goto cleanup_pending;
@@ -11723,14 +11707,24 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 		schedule_work(&work->mmio_work);
 	} else {
-		i915_gem_request_assign(&work->flip_queued_req, request);
+		request = i915_gem_request_alloc(engine, engine->last_context);
+		if (IS_ERR(request)) {
+			ret = PTR_ERR(request);
+			goto cleanup_unpin;
+		}
+
+		ret = i915_gem_object_sync(obj, request);
+		if (ret)
+			goto cleanup_request;
+
 		ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, request,
 						   page_flip_flags);
 		if (ret)
-			goto cleanup_unpin;
+			goto cleanup_request;
 
 		intel_mark_page_flip_active(intel_crtc, work);
 
+		work->flip_queued_req = i915_gem_request_get(request);
 		i915_add_request_no_flush(request);
 	}
 
@@ -11745,11 +11739,11 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 	return 0;
 
+cleanup_request:
+	i915_add_request_no_flush(request);
 cleanup_unpin:
 	intel_unpin_fb_obj(fb, crtc->primary->state->rotation);
 cleanup_pending:
-	if (!IS_ERR_OR_NULL(request))
-		i915_add_request_no_flush(request);
 	atomic_dec(&intel_crtc->unpin_work_count);
 	mutex_unlock(&dev->struct_mutex);
 cleanup:
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 6487794ac96c..a37af1533618 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -655,7 +655,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
 		struct drm_i915_gem_object *obj = vma->obj;
 
 		if (obj->active & other_rings) {
-			ret = i915_gem_object_sync(obj, req->engine, &req);
+			ret = i915_gem_object_sync(obj, req);
 			if (ret)
 				return ret;
 		}
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index 8f1d4d9ef345..651efe4e468e 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -229,11 +229,18 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
 	return 0;
 }
 
+static struct drm_i915_gem_request *alloc_request(struct intel_overlay *overlay)
+{
+	struct drm_i915_private *dev_priv = overlay->i915;
+	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
+
+	return i915_gem_request_alloc(engine, dev_priv->kernel_context);
+}
+
 /* overlay needs to be disable in OCMD reg */
 static int intel_overlay_on(struct intel_overlay *overlay)
 {
 	struct drm_i915_private *dev_priv = overlay->i915;
-	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
 	struct intel_ring *ring;
 	int ret;
@@ -241,7 +248,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 	WARN_ON(overlay->active);
 	WARN_ON(IS_I830(dev_priv) && !(dev_priv->quirks & QUIRK_PIPEA_FORCE));
 
-	req = i915_gem_request_alloc(engine, NULL);
+	req = alloc_request(overlay);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -268,7 +275,6 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 				  bool load_polyphase_filter)
 {
 	struct drm_i915_private *dev_priv = overlay->i915;
-	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
 	struct intel_ring *ring;
 	u32 flip_addr = overlay->flip_addr;
@@ -285,7 +291,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 	if (tmp & (1 << 17))
 		DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp);
 
-	req = i915_gem_request_alloc(engine, NULL);
+	req = alloc_request(overlay);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -338,7 +344,6 @@ static void intel_overlay_off_tail(struct intel_overlay *overlay)
 static int intel_overlay_off(struct intel_overlay *overlay)
 {
 	struct drm_i915_private *dev_priv = overlay->i915;
-	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	struct drm_i915_gem_request *req;
 	struct intel_ring *ring;
 	u32 flip_addr = overlay->flip_addr;
@@ -352,7 +357,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 	 * of the hw. Do it in both cases */
 	flip_addr |= OFC_UPDATE;
 
-	req = i915_gem_request_alloc(engine, NULL);
+	req = alloc_request(overlay);
 	if (IS_ERR(req))
 		return PTR_ERR(req);
 
@@ -412,7 +417,6 @@ static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay)
 static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 {
 	struct drm_i915_private *dev_priv = overlay->i915;
-	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
 	int ret;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
@@ -428,7 +432,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 		struct drm_i915_gem_request *req;
 		struct intel_ring *ring;
 
-		req = i915_gem_request_alloc(engine, NULL);
+		req = alloc_request(overlay);
 		if (IS_ERR(req))
 			return PTR_ERR(req);
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 19/31] drm/i915: Unify legacy/execlists emission of MI_BATCHBUFFER_START

[Chris Wilson]

Both the ->dispatch_execbuffer and ->emit_bb_start callbacks do exactly
the same thing, add MI_BATCHBUFFER_START to the request's ringbuffer -
we need only one vfunc.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_execbuffer.c   |  6 ++--
 drivers/gpu/drm/i915/i915_gem_render_state.c | 16 +++++-----
 drivers/gpu/drm/i915/intel_lrc.c             | 15 ++++++---
 drivers/gpu/drm/i915/intel_ringbuffer.c      | 48 ++++++++++++++--------------
 drivers/gpu/drm/i915/intel_ringbuffer.h      | 12 +++----
 5 files changed, 50 insertions(+), 47 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 75d19932847d..1c0e6f1075b7 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1326,9 +1326,9 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 	if (exec_len == 0)
 		exec_len = params->batch_obj->base.size;
 
-	ret = params->engine->dispatch_execbuffer(params->request,
-						  exec_start, exec_len,
-						  params->dispatch_flags);
+	ret = params->engine->emit_bb_start(params->request,
+					    exec_start, exec_len,
+					    params->dispatch_flags);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index b2be4676a5cf..2ba759f3ab6f 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -234,18 +234,18 @@ int i915_gem_render_state_init(struct drm_i915_gem_request *req)
 	if (so.rodata == NULL)
 		return 0;
 
-	ret = req->engine->dispatch_execbuffer(req, so.ggtt_offset,
-					     so.rodata->batch_items * 4,
-					     I915_DISPATCH_SECURE);
+	ret = req->engine->emit_bb_start(req, so.ggtt_offset,
+					 so.rodata->batch_items * 4,
+					 I915_DISPATCH_SECURE);
 	if (ret)
 		goto out;
 
 	if (so.aux_batch_size > 8) {
-		ret = req->engine->dispatch_execbuffer(req,
-						     (so.ggtt_offset +
-						      so.aux_batch_offset),
-						     so.aux_batch_size,
-						     I915_DISPATCH_SECURE);
+		ret = req->engine->emit_bb_start(req,
+						 (so.ggtt_offset +
+						  so.aux_batch_offset),
+						 so.aux_batch_size,
+						 I915_DISPATCH_SECURE);
 		if (ret)
 			goto out;
 	}
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index a37af1533618..567d94de3300 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -859,7 +859,9 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	exec_start = params->batch_obj_vm_offset +
 		     args->batch_start_offset;
 
-	ret = engine->emit_bb_start(params->request, exec_start, params->dispatch_flags);
+	ret = engine->emit_bb_start(params->request,
+				    exec_start, args->batch_len,
+				    params->dispatch_flags);
 	if (ret)
 		return ret;
 
@@ -1543,7 +1545,8 @@ static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
 }
 
 static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
-			      u64 offset, unsigned dispatch_flags)
+			      u64 offset, u32 len,
+			      unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	bool ppgtt = !(dispatch_flags & I915_DISPATCH_SECURE);
@@ -1819,13 +1822,15 @@ static int intel_lr_context_render_state_init(struct drm_i915_gem_request *req)
 		return 0;
 
 	ret = req->engine->emit_bb_start(req, so.ggtt_offset,
-				       I915_DISPATCH_SECURE);
+					 so.rodata->batch_items * 4,
+					 I915_DISPATCH_SECURE);
 	if (ret)
 		goto out;
 
 	ret = req->engine->emit_bb_start(req,
-				       (so.ggtt_offset + so.aux_batch_offset),
-				       I915_DISPATCH_SECURE);
+					 (so.ggtt_offset + so.aux_batch_offset),
+					 so.aux_batch_size,
+					 I915_DISPATCH_SECURE);
 	if (ret)
 		goto out;
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index bd3ed9e12afe..0a2e25e67450 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1794,9 +1794,9 @@ gen8_irq_disable(struct intel_engine_cs *engine)
 }
 
 static int
-i965_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			 u64 offset, u32 length,
-			 unsigned dispatch_flags)
+i965_emit_bb_start(struct drm_i915_gem_request *req,
+		   u64 offset, u32 length,
+		   unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -1821,9 +1821,9 @@ i965_dispatch_execbuffer(struct drm_i915_gem_request *req,
 #define I830_TLB_ENTRIES (2)
 #define I830_WA_SIZE max(I830_TLB_ENTRIES*4096, I830_BATCH_LIMIT)
 static int
-i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			 u64 offset, u32 len,
-			 unsigned dispatch_flags)
+i830_emit_bb_start(struct drm_i915_gem_request *req,
+		   u64 offset, u32 len,
+		   unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	u32 cs_offset = req->engine->scratch.gtt_offset;
@@ -1883,9 +1883,9 @@ i830_dispatch_execbuffer(struct drm_i915_gem_request *req,
 }
 
 static int
-i915_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			 u64 offset, u32 len,
-			 unsigned dispatch_flags)
+i915_emit_bb_start(struct drm_i915_gem_request *req,
+		   u64 offset, u32 len,
+		   unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -2578,9 +2578,9 @@ static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 }
 
 static int
-gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			      u64 offset, u32 len,
-			      unsigned dispatch_flags)
+gen8_emit_bb_start(struct drm_i915_gem_request *req,
+		   u64 offset, u32 len,
+		   unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	bool ppgtt = USES_PPGTT(req->i915) &&
@@ -2604,9 +2604,9 @@ gen8_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 }
 
 static int
-hsw_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			     u64 offset, u32 len,
-			     unsigned dispatch_flags)
+hsw_emit_bb_start(struct drm_i915_gem_request *req,
+		  u64 offset, u32 len,
+		  unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -2629,9 +2629,9 @@ hsw_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 }
 
 static int
-gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
-			      u64 offset, u32 len,
-			      unsigned dispatch_flags)
+gen6_emit_bb_start(struct drm_i915_gem_request *req,
+		   u64 offset, u32 len,
+		   unsigned int dispatch_flags)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -2835,15 +2835,15 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 		engine->add_request = gen6_add_request;
 
 	if (INTEL_GEN(dev_priv) >= 8)
-		engine->dispatch_execbuffer = gen8_ring_dispatch_execbuffer;
+		engine->emit_bb_start = gen8_emit_bb_start;
 	else if (INTEL_GEN(dev_priv) >= 6)
-		engine->dispatch_execbuffer = gen6_ring_dispatch_execbuffer;
+		engine->emit_bb_start = gen6_emit_bb_start;
 	else if (INTEL_GEN(dev_priv) >= 4)
-		engine->dispatch_execbuffer = i965_dispatch_execbuffer;
+		engine->emit_bb_start = i965_emit_bb_start;
 	else if (IS_I830(dev_priv) || IS_845G(dev_priv))
-		engine->dispatch_execbuffer = i830_dispatch_execbuffer;
+		engine->emit_bb_start = i830_emit_bb_start;
 	else
-		engine->dispatch_execbuffer = i915_dispatch_execbuffer;
+		engine->emit_bb_start = i915_emit_bb_start;
 
 	intel_ring_init_irq(dev_priv, engine);
 	intel_ring_init_semaphores(dev_priv, engine);
@@ -2881,7 +2881,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
 	}
 
 	if (IS_HASWELL(dev_priv))
-		engine->dispatch_execbuffer = hsw_ring_dispatch_execbuffer;
+		engine->emit_bb_start = hsw_emit_bb_start;
 
 	engine->init_hw = init_render_ring;
 	engine->cleanup = render_ring_cleanup;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 05ac78a7ea64..cbd46d5e8d6b 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -214,12 +214,6 @@ struct intel_engine_cs {
 	 * monotonic, even if not coherent.
 	 */
 	void		(*irq_seqno_barrier)(struct intel_engine_cs *engine);
-	int		(*dispatch_execbuffer)(struct drm_i915_gem_request *req,
-					       u64 offset, u32 length,
-					       unsigned dispatch_flags);
-#define I915_DISPATCH_SECURE 0x1
-#define I915_DISPATCH_PINNED 0x2
-#define I915_DISPATCH_RS     0x4
 	void		(*cleanup)(struct intel_engine_cs *engine);
 
 	/* GEN8 signal/wait table - never trust comments!
@@ -295,7 +289,11 @@ struct intel_engine_cs {
 				      u32 invalidate_domains,
 				      u32 flush_domains);
 	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
-					 u64 offset, unsigned dispatch_flags);
+					 u64 offset, u32 length,
+					 unsigned int dispatch_flags);
+#define I915_DISPATCH_SECURE 0x1
+#define I915_DISPATCH_PINNED 0x2
+#define I915_DISPATCH_RS     0x4
 
 	/**
 	 * List of objects currently involved in rendering from the
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 20/31] drm/i915: Remove intel_ring_get_tail()

[Chris Wilson]

Joonas doesn't like the tiny function, especially if I go around making
it more complicated and using it elsewhere. To remove that temptation,
remove the function!

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_request.c | 8 ++++----
 drivers/gpu/drm/i915/intel_ringbuffer.h | 5 -----
 2 files changed, 4 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 995ae99e9cf7..1c185e293bf0 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -419,7 +419,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 * should already have been reserved in the ring buffer. Let the ring
 	 * know that it is time to use that space up.
 	 */
-	request_start = intel_ring_get_tail(ring);
+	request_start = ring->tail;
 	reserved_tail = request->reserved_space;
 	request->reserved_space = 0;
 
@@ -465,19 +465,19 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 * GPU processing the request, we never over-estimate the
 	 * position of the head.
 	 */
-	request->postfix = intel_ring_get_tail(ring);
+	request->postfix = ring->tail;
 
 	if (i915.enable_execlists) {
 		ret = engine->emit_request(request);
 	} else {
 		ret = engine->add_request(request);
 
-		request->tail = intel_ring_get_tail(ring);
+		request->tail = ring->tail;
 	}
 	/* Not allowed to fail! */
 	WARN(ret, "emit|add_request failed: %d!\n", ret);
 	/* Sanity check that the reserved size was large enough. */
-	ret = intel_ring_get_tail(ring) - request_start;
+	ret = ring->tail - request_start;
 	if (ret < 0)
 		ret += ring->size;
 	WARN_ONCE(ret > reserved_tail,
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index cbd46d5e8d6b..13b816f2d264 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -488,11 +488,6 @@ static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine)
 
 int init_workarounds_ring(struct intel_engine_cs *engine);
 
-static inline u32 intel_ring_get_tail(struct intel_ring *ring)
-{
-	return ring->tail;
-}
-
 /*
  * Arbitrary size for largest possible 'add request' sequence. The code paths
  * are complex and variable. Empirical measurement shows that the worst case
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

If we rewrite the I915_WRITE_TAIL specialisation for the legacy
ringbuffer as submitting the request onto the ringbuffer, we can unify
the vfunc with both execlists and GuC in the next patch.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_request.c |  8 ++---
 drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c | 53 +++++++++++++++++----------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  3 +-
 4 files changed, 32 insertions(+), 34 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 1c185e293bf0..8814e9c5266b 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -467,15 +467,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 */
 	request->postfix = ring->tail;
 
-	if (i915.enable_execlists) {
+	if (i915.enable_execlists)
 		ret = engine->emit_request(request);
-	} else {
+	else
 		ret = engine->add_request(request);
-
-		request->tail = ring->tail;
-	}
 	/* Not allowed to fail! */
 	WARN(ret, "emit|add_request failed: %d!\n", ret);
+
 	/* Sanity check that the reserved size was large enough. */
 	ret = ring->tail - request_start;
 	if (ret < 0)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 567d94de3300..250edb2bcef7 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
 	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
 	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
 
-	reg_state[CTX_RING_TAIL+1] = rq->tail;
+	reg_state[CTX_RING_TAIL+1] = rq->tail % (rq->ring->size - 1);
 
 	/* True 32b PPGTT with dynamic page allocation: update PDP
 	 * registers and point the unallocated PDPs to scratch page.
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 0a2e25e67450..3e1049c972e0 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -58,14 +58,6 @@ void intel_ring_update_space(struct intel_ring *ring)
 					 ring->tail, ring->size);
 }
 
-static void __intel_engine_submit(struct intel_engine_cs *engine)
-{
-	struct intel_ring *ring = engine->buffer;
-
-	ring->tail &= ring->size - 1;
-	engine->write_tail(engine, ring->tail);
-}
-
 static int
 gen2_render_ring_flush(struct drm_i915_gem_request *req,
 		       u32	invalidate_domains,
@@ -421,13 +413,6 @@ gen8_render_ring_flush(struct drm_i915_gem_request *req,
 	return gen8_emit_pipe_control(req, flags, scratch_addr);
 }
 
-static void ring_write_tail(struct intel_engine_cs *engine,
-			    u32 value)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-	I915_WRITE_TAIL(engine, value);
-}
-
 u64 intel_engine_get_active_head(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
@@ -541,7 +526,7 @@ static bool stop_ring(struct intel_engine_cs *engine)
 
 	I915_WRITE_CTL(engine, 0);
 	I915_WRITE_HEAD(engine, 0);
-	engine->write_tail(engine, 0);
+	I915_WRITE_TAIL(engine, 0);
 
 	if (!IS_GEN2(dev_priv)) {
 		(void)I915_READ_CTL(engine);
@@ -1482,7 +1467,10 @@ gen6_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
 	intel_ring_emit(ring, req->fence.seqno);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
-	__intel_engine_submit(engine);
+	intel_ring_advance(ring);
+
+	req->tail = ring->tail;
+	engine->submit_request(req);
 
 	return 0;
 }
@@ -1512,7 +1500,9 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, 0);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_emit(ring, MI_NOOP);
-	__intel_engine_submit(engine);
+
+	req->tail = ring->tail;
+	engine->submit_request(req);
 
 	return 0;
 }
@@ -1731,11 +1721,22 @@ i9xx_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
 	intel_ring_emit(ring, req->fence.seqno);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
-	__intel_engine_submit(req->engine);
+	intel_ring_advance(ring);
+
+	req->tail = ring->tail;
+	req->engine->submit_request(req);
 
 	return 0;
 }
 
+static void i9xx_submit_request(struct drm_i915_gem_request *request)
+{
+	struct drm_i915_private *dev_priv = request->i915;
+
+	I915_WRITE_TAIL(request->engine,
+			request->tail % (request->ring->size - 1));
+}
+
 static void
 gen6_irq_enable(struct intel_engine_cs *engine)
 {
@@ -2494,10 +2495,9 @@ void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
 	rcu_read_unlock();
 }
 
-static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
-				     u32 value)
+static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
+	struct drm_i915_private *dev_priv = request->i915;
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
@@ -2521,8 +2521,9 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
 
 	/* Now that the ring is fully powered up, update the tail */
-	I915_WRITE_FW(RING_TAIL(engine->mmio_base), value);
-	POSTING_READ_FW(RING_TAIL(engine->mmio_base));
+	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
+		      request->tail % (request->ring->size - 1));
+	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
 
 	/* Let the ring send IDLE messages to the GT again,
 	 * and so let it sleep to conserve power when idle.
@@ -2828,7 +2829,7 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 				      struct intel_engine_cs *engine)
 {
 	engine->init_hw = init_ring_common;
-	engine->write_tail = ring_write_tail;
+	engine->submit_request = i9xx_submit_request;
 
 	engine->add_request = i9xx_add_request;
 	if (INTEL_GEN(dev_priv) >= 6)
@@ -2912,7 +2913,7 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 	if (INTEL_GEN(dev_priv) >= 6) {
 		/* gen6 bsd needs a special wa for tail updates */
 		if (IS_GEN6(dev_priv))
-			engine->write_tail = gen6_bsd_ring_write_tail;
+			engine->submit_request = gen6_bsd_submit_request;
 		engine->emit_flush = gen6_bsd_ring_flush;
 		if (INTEL_GEN(dev_priv) < 8)
 			engine->irq_enable_mask = GT_BSD_USER_INTERRUPT;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 13b816f2d264..5428a3c288d5 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -204,8 +204,6 @@ struct intel_engine_cs {
 
 	int		(*init_context)(struct drm_i915_gem_request *req);
 
-	void		(*write_tail)(struct intel_engine_cs *engine,
-				      u32 value);
 	int		(*add_request)(struct drm_i915_gem_request *req);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
@@ -294,6 +292,7 @@ struct intel_engine_cs {
 #define I915_DISPATCH_SECURE 0x1
 #define I915_DISPATCH_PINNED 0x2
 #define I915_DISPATCH_RS     0x4
+	void		(*submit_request)(struct drm_i915_gem_request *req);
 
 	/**
 	 * List of objects currently involved in rendering from the
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 22/31] drm/i915: Unify request submission

[Chris Wilson]

Move request submission from emit_request into its own common vfunc
from i915_add_request().

v2: Convert I915_DISPATCH_flags to BIT(x) whilst passing
v3: Rename a few functions to match.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_request.c    |  8 +++-----
 drivers/gpu/drm/i915/i915_guc_submission.c |  9 ++++++---
 drivers/gpu/drm/i915/intel_guc.h           |  1 -
 drivers/gpu/drm/i915/intel_lrc.c           | 18 +++++++-----------
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 23 +++++++++--------------
 drivers/gpu/drm/i915/intel_ringbuffer.h    | 23 +++++++++++------------
 6 files changed, 36 insertions(+), 46 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 8814e9c5266b..f4bf9f669eed 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -467,12 +467,9 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 */
 	request->postfix = ring->tail;
 
-	if (i915.enable_execlists)
-		ret = engine->emit_request(request);
-	else
-		ret = engine->add_request(request);
 	/* Not allowed to fail! */
-	WARN(ret, "emit|add_request failed: %d!\n", ret);
+	ret = engine->emit_request(request);
+	WARN(ret, "(%s)->emit_request failed: %d!\n", engine->name, ret);
 
 	/* Sanity check that the reserved size was large enough. */
 	ret = ring->tail - request_start;
@@ -484,6 +481,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 		  reserved_tail, ret);
 
 	i915_gem_mark_busy(engine);
+	engine->submit_request(request);
 }
 
 static unsigned long local_clock_us(unsigned int *cpu)
diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index eccd34832fe6..32d0e1890950 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -585,7 +585,7 @@ static int guc_ring_doorbell(struct i915_guc_client *gc)
  * The only error here arises if the doorbell hardware isn't functioning
  * as expected, which really shouln't happen.
  */
-int i915_guc_submit(struct drm_i915_gem_request *rq)
+static void i915_guc_submit(struct drm_i915_gem_request *rq)
 {
 	unsigned int engine_id = rq->engine->id;
 	struct intel_guc *guc = &rq->i915->guc;
@@ -602,8 +602,6 @@ int i915_guc_submit(struct drm_i915_gem_request *rq)
 
 	guc->submissions[engine_id] += 1;
 	guc->last_seqno[engine_id] = rq->fence.seqno;
-
-	return b_ret;
 }
 
 /*
@@ -992,6 +990,7 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
 {
 	struct intel_guc *guc = &dev_priv->guc;
 	struct i915_guc_client *client;
+	struct intel_engine_cs *engine;
 
 	/* client for execbuf submission */
 	client = guc_client_alloc(dev_priv,
@@ -1006,6 +1005,10 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
 	host2guc_sample_forcewake(guc, client);
 	guc_init_doorbell_hw(guc);
 
+	/* Take over from manual control of ELSP (execlists) */
+	for_each_engine(engine, dev_priv)
+		engine->submit_request = i915_guc_submit;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/intel_guc.h b/drivers/gpu/drm/i915/intel_guc.h
index 3e3e743740c0..623cf26cd784 100644
--- a/drivers/gpu/drm/i915/intel_guc.h
+++ b/drivers/gpu/drm/i915/intel_guc.h
@@ -160,7 +160,6 @@ extern int intel_guc_resume(struct drm_device *dev);
 int i915_guc_submission_init(struct drm_i915_private *dev_priv);
 int i915_guc_submission_enable(struct drm_i915_private *dev_priv);
 int i915_guc_wq_check_space(struct drm_i915_gem_request *rq);
-int i915_guc_submit(struct drm_i915_gem_request *rq);
 void i915_guc_submission_disable(struct drm_i915_private *dev_priv);
 void i915_guc_submission_fini(struct drm_i915_private *dev_priv);
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 250edb2bcef7..a9ca31c113c3 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -738,7 +738,7 @@ err_unpin:
 }
 
 /*
- * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
+ * intel_logical_ring_advance() - advance the tail and prepare for submission
  * @request: Request to advance the logical ringbuffer of.
  *
  * The tail is updated in our logical ringbuffer struct, not in the actual context. What
@@ -747,7 +747,7 @@ err_unpin:
  * point, the tail *inside* the context is updated and the ELSP written to.
  */
 static int
-intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
+intel_logical_ring_advance(struct drm_i915_gem_request *request)
 {
 	struct intel_ring *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
@@ -773,12 +773,6 @@ intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
 	 */
 	request->previous_context = engine->last_context;
 	engine->last_context = request->ctx;
-
-	if (i915.enable_guc_submission)
-		i915_guc_submit(request);
-	else
-		execlists_context_queue(request);
-
 	return 0;
 }
 
@@ -1775,7 +1769,7 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)
 	intel_ring_emit(ring, request->fence.seqno);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_emit(ring, MI_NOOP);
-	return intel_logical_ring_advance_and_submit(request);
+	return intel_logical_ring_advance(request);
 }
 
 static int gen8_emit_request_render(struct drm_i915_gem_request *request)
@@ -1806,7 +1800,7 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 	intel_ring_emit(ring, 0);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_emit(ring, MI_NOOP);
-	return intel_logical_ring_advance_and_submit(request);
+	return intel_logical_ring_advance(request);
 }
 
 static int intel_lr_context_render_state_init(struct drm_i915_gem_request *req)
@@ -1912,8 +1906,10 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 {
 	/* Default vfuncs which can be overriden by each engine. */
 	engine->init_hw = gen8_init_common_ring;
-	engine->emit_request = gen8_emit_request;
 	engine->emit_flush = gen8_emit_flush;
+	engine->emit_request = gen8_emit_request;
+	engine->submit_request = execlists_context_queue;
+
 	engine->irq_enable = gen8_logical_ring_enable_irq;
 	engine->irq_disable = gen8_logical_ring_disable_irq;
 	engine->emit_bb_start = gen8_emit_bb_start;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 3e1049c972e0..2fa7db5331c3 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1441,15 +1441,14 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 }
 
 /**
- * gen6_add_request - Update the semaphore mailbox registers
+ * gen6_emit_request - Update the semaphore mailbox registers
  *
  * @request - request to write to the ring
  *
  * Update the mailbox registers in the *other* rings with the current seqno.
  * This acts like a signal in the canonical semaphore.
  */
-static int
-gen6_add_request(struct drm_i915_gem_request *req)
+static int gen6_emit_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
 	struct intel_ring *ring = req->ring;
@@ -1470,13 +1469,11 @@ gen6_add_request(struct drm_i915_gem_request *req)
 	intel_ring_advance(ring);
 
 	req->tail = ring->tail;
-	engine->submit_request(req);
 
 	return 0;
 }
 
-static int
-gen8_render_add_request(struct drm_i915_gem_request *req)
+static int gen8_render_emit_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
 	struct intel_ring *ring = req->ring;
@@ -1500,9 +1497,9 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, 0);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	req->tail = ring->tail;
-	engine->submit_request(req);
 
 	return 0;
 }
@@ -1707,8 +1704,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 	return 0;
 }
 
-static int
-i9xx_add_request(struct drm_i915_gem_request *req)
+static int i9xx_emit_request(struct drm_i915_gem_request *req)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -1724,7 +1720,6 @@ i9xx_add_request(struct drm_i915_gem_request *req)
 	intel_ring_advance(ring);
 
 	req->tail = ring->tail;
-	req->engine->submit_request(req);
 
 	return 0;
 }
@@ -2829,11 +2824,11 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 				      struct intel_engine_cs *engine)
 {
 	engine->init_hw = init_ring_common;
-	engine->submit_request = i9xx_submit_request;
 
-	engine->add_request = i9xx_add_request;
+	engine->emit_request = i9xx_emit_request;
 	if (INTEL_GEN(dev_priv) >= 6)
-		engine->add_request = gen6_add_request;
+		engine->emit_request = gen6_emit_request;
+	engine->submit_request = i9xx_submit_request;
 
 	if (INTEL_GEN(dev_priv) >= 8)
 		engine->emit_bb_start = gen8_emit_bb_start;
@@ -2862,7 +2857,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
 
 	if (INTEL_GEN(dev_priv) >= 8) {
 		engine->init_context = intel_rcs_ctx_init;
-		engine->add_request = gen8_render_add_request;
+		engine->emit_request = gen8_render_emit_request;
 		engine->emit_flush = gen8_render_ring_flush;
 		if (i915.semaphores)
 			engine->semaphore.signal = gen8_rcs_signal;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 5428a3c288d5..fdf085495e3a 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -204,7 +204,17 @@ struct intel_engine_cs {
 
 	int		(*init_context)(struct drm_i915_gem_request *req);
 
-	int		(*add_request)(struct drm_i915_gem_request *req);
+	int		(*emit_flush)(struct drm_i915_gem_request *request,
+				      u32 invalidate_domains,
+				      u32 flush_domains);
+	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
+					 u64 offset, u32 length,
+					 unsigned int dispatch_flags);
+#define I915_DISPATCH_SECURE BIT(0)
+#define I915_DISPATCH_PINNED BIT(1)
+#define I915_DISPATCH_RS     BIT(2)
+	int		(*emit_request)(struct drm_i915_gem_request *req);
+	void		(*submit_request)(struct drm_i915_gem_request *req);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
 	 * However, the up-to-date seqno is not always required and the last
@@ -282,17 +292,6 @@ struct intel_engine_cs {
 	unsigned int idle_lite_restore_wa;
 	bool disable_lite_restore_wa;
 	u32 ctx_desc_template;
-	int		(*emit_request)(struct drm_i915_gem_request *request);
-	int		(*emit_flush)(struct drm_i915_gem_request *request,
-				      u32 invalidate_domains,
-				      u32 flush_domains);
-	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
-					 u64 offset, u32 length,
-					 unsigned int dispatch_flags);
-#define I915_DISPATCH_SECURE 0x1
-#define I915_DISPATCH_PINNED 0x2
-#define I915_DISPATCH_RS     0x4
-	void		(*submit_request)(struct drm_i915_gem_request *req);
 
 	/**
 	 * List of objects currently involved in rendering from the
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 23/31] drm/i915/lrc: Update function names to match request flow

[Chris Wilson]

With adding engine->submit_request, we now have a bunch of functions
with similar names used at different stages of the execlist submission.
Try a different coat of paint, to hopefully reduce confusion between the
requests, intel_engine_cs and the actual execlists submision process.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/intel_lrc.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index a9ca31c113c3..149a0dc7aeed 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -384,8 +384,8 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
 		execlists_update_context_pdps(ppgtt, reg_state);
 }
 
-static void execlists_submit_requests(struct drm_i915_gem_request *rq0,
-				      struct drm_i915_gem_request *rq1)
+static void execlists_elsp_submit_contexts(struct drm_i915_gem_request *rq0,
+					   struct drm_i915_gem_request *rq1)
 {
 	struct drm_i915_private *dev_priv = rq0->i915;
 	unsigned int fw_domains = rq0->engine->fw_domains;
@@ -418,7 +418,7 @@ static inline void execlists_context_status_change(
 	atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq);
 }
 
-static void execlists_context_unqueue(struct intel_engine_cs *engine)
+static void execlists_unqueue(struct intel_engine_cs *engine)
 {
 	struct drm_i915_gem_request *req0 = NULL, *req1 = NULL;
 	struct drm_i915_gem_request *cursor, *tmp;
@@ -486,7 +486,7 @@ static void execlists_context_unqueue(struct intel_engine_cs *engine)
 		req0->tail &= req0->ring->size - 1;
 	}
 
-	execlists_submit_requests(req0, req1);
+	execlists_elsp_submit_contexts(req0, req1);
 }
 
 static unsigned int
@@ -597,7 +597,7 @@ static void intel_lrc_irq_handler(unsigned long data)
 	if (submit_contexts) {
 		if (!engine->disable_lite_restore_wa ||
 		    (csb[i][0] & GEN8_CTX_STATUS_ACTIVE_IDLE))
-			execlists_context_unqueue(engine);
+			execlists_unqueue(engine);
 	}
 
 	spin_unlock(&engine->execlist_lock);
@@ -606,7 +606,7 @@ static void intel_lrc_irq_handler(unsigned long data)
 		DRM_ERROR("More than two context complete events?\n");
 }
 
-static void execlists_context_queue(struct drm_i915_gem_request *request)
+static void execlists_submit_request(struct drm_i915_gem_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct drm_i915_gem_request *cursor;
@@ -637,7 +637,7 @@ static void execlists_context_queue(struct drm_i915_gem_request *request)
 	list_add_tail(&request->execlist_link, &engine->execlist_queue);
 	request->ctx_hw_id = request->ctx->hw_id;
 	if (num_elements == 0)
-		execlists_context_unqueue(engine);
+		execlists_unqueue(engine);
 
 	spin_unlock_bh(&engine->execlist_lock);
 }
@@ -1908,7 +1908,7 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 	engine->init_hw = gen8_init_common_ring;
 	engine->emit_flush = gen8_emit_flush;
 	engine->emit_request = gen8_emit_request;
-	engine->submit_request = execlists_context_queue;
+	engine->submit_request = execlists_submit_request;
 
 	engine->irq_enable = gen8_logical_ring_enable_irq;
 	engine->irq_disable = gen8_logical_ring_disable_irq;
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 24/31] drm/i915: Stop passing caller's num_dwords to engine->semaphore.signal()

[Chris Wilson]

Rather than pass in the num_dwords that the caller wishes to use after
the signal command packet, split the breadcrumb emission into two phases
and have both the signal and breadcrumb individiually acquire space on
the ring. This makes the interface simpler for the reader, and will
simplify for patches.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 51 ++++++++++++++-------------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  4 +--
 2 files changed, 23 insertions(+), 32 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 2fa7db5331c3..68df689a9d1e 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1322,10 +1322,8 @@ static void render_ring_cleanup(struct intel_engine_cs *engine)
 	intel_fini_pipe_control(engine);
 }
 
-static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
-			   unsigned int num_dwords)
+static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req)
 {
-#define MBOX_UPDATE_DWORDS 8
 	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
@@ -1333,10 +1331,7 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 	int ret, num_rings;
 
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	num_dwords += (num_rings-1) * MBOX_UPDATE_DWORDS;
-#undef MBOX_UPDATE_DWORDS
-
-	ret = intel_ring_begin(signaller_req, num_dwords);
+	ret = intel_ring_begin(signaller_req, (num_rings-1) * 8);
 	if (ret)
 		return ret;
 
@@ -1360,14 +1355,13 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
 				MI_SEMAPHORE_TARGET(waiter->hw_id));
 		intel_ring_emit(signaller, 0);
 	}
+	intel_ring_advance(signaller);
 
 	return 0;
 }
 
-static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
-			   unsigned int num_dwords)
+static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req)
 {
-#define MBOX_UPDATE_DWORDS 6
 	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
 	struct intel_engine_cs *waiter;
@@ -1375,10 +1369,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 	int ret, num_rings;
 
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	num_dwords += (num_rings-1) * MBOX_UPDATE_DWORDS;
-#undef MBOX_UPDATE_DWORDS
-
-	ret = intel_ring_begin(signaller_req, num_dwords);
+	ret = intel_ring_begin(signaller_req, (num_rings-1) * 6);
 	if (ret)
 		return ret;
 
@@ -1400,12 +1391,12 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
 				MI_SEMAPHORE_TARGET(waiter->hw_id));
 		intel_ring_emit(signaller, 0);
 	}
+	intel_ring_advance(signaller);
 
 	return 0;
 }
 
-static int gen6_signal(struct drm_i915_gem_request *signaller_req,
-		       unsigned int num_dwords)
+static int gen6_signal(struct drm_i915_gem_request *signaller_req)
 {
 	struct intel_ring *signaller = signaller_req->ring;
 	struct drm_i915_private *dev_priv = signaller_req->i915;
@@ -1413,12 +1404,8 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 	enum intel_engine_id id;
 	int ret, num_rings;
 
-#define MBOX_UPDATE_DWORDS 3
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	num_dwords += round_up((num_rings-1) * MBOX_UPDATE_DWORDS, 2);
-#undef MBOX_UPDATE_DWORDS
-
-	ret = intel_ring_begin(signaller_req, num_dwords);
+	ret = intel_ring_begin(signaller_req, round_up((num_rings-1) * 3, 2));
 	if (ret)
 		return ret;
 
@@ -1436,6 +1423,7 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
 	/* If num_dwords was rounded, make sure the tail pointer is correct */
 	if (num_rings % 2 == 0)
 		intel_ring_emit(signaller, MI_NOOP);
+	intel_ring_advance(signaller);
 
 	return 0;
 }
@@ -1454,11 +1442,13 @@ static int gen6_emit_request(struct drm_i915_gem_request *req)
 	struct intel_ring *ring = req->ring;
 	int ret;
 
-	if (engine->semaphore.signal)
-		ret = engine->semaphore.signal(req, 4);
-	else
-		ret = intel_ring_begin(req, 4);
+	if (engine->semaphore.signal) {
+		ret = engine->semaphore.signal(req);
+		if (ret)
+			return ret;
+	}
 
+	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
 
@@ -1479,10 +1469,13 @@ static int gen8_render_emit_request(struct drm_i915_gem_request *req)
 	struct intel_ring *ring = req->ring;
 	int ret;
 
-	if (engine->semaphore.signal)
-		ret = engine->semaphore.signal(req, 8);
-	else
-		ret = intel_ring_begin(req, 8);
+	if (engine->semaphore.signal) {
+		ret = engine->semaphore.signal(req);
+		if (ret)
+			return ret;
+	}
+
+	ret = intel_ring_begin(req, 8);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index fdf085495e3a..a137fd218c83 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -278,9 +278,7 @@ struct intel_engine_cs {
 		int	(*sync_to)(struct drm_i915_gem_request *to_req,
 				   struct intel_engine_cs *from,
 				   u32 seqno);
-		int	(*signal)(struct drm_i915_gem_request *signaller_req,
-				  /* num_dwords needed by caller */
-				  unsigned int num_dwords);
+		int	(*signal)(struct drm_i915_gem_request *signaller_req);
 	} semaphore;
 
 	/* Execlists */
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 25/31] drm/i915: Reuse legacy breadcrumbs + tail emission

[Chris Wilson]

As GEN6+ is now a simple variant on the basic breadcrumbs + tail write,
reuse the common code.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 74 +++++++++++++--------------------
 1 file changed, 30 insertions(+), 44 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 68df689a9d1e..a42d878009e6 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1428,26 +1428,19 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req)
 	return 0;
 }
 
-/**
- * gen6_emit_request - Update the semaphore mailbox registers
- *
- * @request - request to write to the ring
- *
- * Update the mailbox registers in the *other* rings with the current seqno.
- * This acts like a signal in the canonical semaphore.
- */
-static int gen6_emit_request(struct drm_i915_gem_request *req)
+static void i9xx_submit_request(struct drm_i915_gem_request *request)
+{
+	struct drm_i915_private *dev_priv = request->i915;
+
+	I915_WRITE_TAIL(request->engine,
+			request->tail % (request->ring->size - 1));
+}
+
+static int i9xx_emit_request(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *engine = req->engine;
 	struct intel_ring *ring = req->ring;
 	int ret;
 
-	if (engine->semaphore.signal) {
-		ret = engine->semaphore.signal(req);
-		if (ret)
-			return ret;
-	}
-
 	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
@@ -1463,6 +1456,27 @@ static int gen6_emit_request(struct drm_i915_gem_request *req)
 	return 0;
 }
 
+/**
+ * gen6_emit_request - Update the semaphore mailbox registers
+ *
+ * @request - request to write to the ring
+ *
+ * Update the mailbox registers in the *other* rings with the current seqno.
+ * This acts like a signal in the canonical semaphore.
+ */
+static int gen6_emit_request(struct drm_i915_gem_request *req)
+{
+	if (req->engine->semaphore.signal) {
+		int ret;
+
+		ret = req->engine->semaphore.signal(req);
+		if (ret)
+			return ret;
+	}
+
+	return i9xx_emit_request(req);
+}
+
 static int gen8_render_emit_request(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
@@ -1697,34 +1711,6 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
 	return 0;
 }
 
-static int i9xx_emit_request(struct drm_i915_gem_request *req)
-{
-	struct intel_ring *ring = req->ring;
-	int ret;
-
-	ret = intel_ring_begin(req, 4);
-	if (ret)
-		return ret;
-
-	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
-	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
-	intel_ring_emit(ring, req->fence.seqno);
-	intel_ring_emit(ring, MI_USER_INTERRUPT);
-	intel_ring_advance(ring);
-
-	req->tail = ring->tail;
-
-	return 0;
-}
-
-static void i9xx_submit_request(struct drm_i915_gem_request *request)
-{
-	struct drm_i915_private *dev_priv = request->i915;
-
-	I915_WRITE_TAIL(request->engine,
-			request->tail % (request->ring->size - 1));
-}
-
 static void
 gen6_irq_enable(struct intel_engine_cs *engine)
 {
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 26/31] drm/i915/ringbuffer: Specialise SNB+ request emission for semaphores

[Chris Wilson]

As gen6_emit_request() only differs from i9xx_emit_request() when
semaphores are enabled, only use the specialised vfunc in that scenario.

v2: Reorder semaphore init so as to keep engine->emit_request default
vfunc selection compact.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 24 +++++++++++-------------
 1 file changed, 11 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index a42d878009e6..0fe4b000749a 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1457,22 +1457,20 @@ static int i9xx_emit_request(struct drm_i915_gem_request *req)
 }
 
 /**
- * gen6_emit_request - Update the semaphore mailbox registers
+ * gen6_sema_emit_request - Update the semaphore mailbox registers
  *
  * @request - request to write to the ring
  *
  * Update the mailbox registers in the *other* rings with the current seqno.
  * This acts like a signal in the canonical semaphore.
  */
-static int gen6_emit_request(struct drm_i915_gem_request *req)
+static int gen6_sema_emit_request(struct drm_i915_gem_request *req)
 {
-	if (req->engine->semaphore.signal) {
-		int ret;
+	int ret;
 
-		ret = req->engine->semaphore.signal(req);
-		if (ret)
-			return ret;
-	}
+	ret = req->engine->semaphore.signal(req);
+	if (ret)
+		return ret;
 
 	return i9xx_emit_request(req);
 }
@@ -2802,12 +2800,15 @@ static void intel_ring_init_irq(struct drm_i915_private *dev_priv,
 static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 				      struct intel_engine_cs *engine)
 {
+	intel_ring_init_irq(dev_priv, engine);
+	intel_ring_init_semaphores(dev_priv, engine);
+
 	engine->init_hw = init_ring_common;
 
 	engine->emit_request = i9xx_emit_request;
-	if (INTEL_GEN(dev_priv) >= 6)
-		engine->emit_request = gen6_emit_request;
 	engine->submit_request = i9xx_submit_request;
+	if (i915.semaphores)
+		engine->emit_request = gen6_sema_emit_request;
 
 	if (INTEL_GEN(dev_priv) >= 8)
 		engine->emit_bb_start = gen8_emit_bb_start;
@@ -2819,9 +2820,6 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 		engine->emit_bb_start = i830_emit_bb_start;
 	else
 		engine->emit_bb_start = i915_emit_bb_start;
-
-	intel_ring_init_irq(dev_priv, engine);
-	intel_ring_init_semaphores(dev_priv, engine);
 }
 
 int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 27/31] drm/i915: Remove duplicate golden render state init from execlists

[Chris Wilson]

Now that we use the same vfuncs for emitting the batch buffer in both
execlists and legacy, the golden render state initialisation is
identical between both.

v2: gcc wants so.ggtt_offset initialised (even though it is not used)

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_render_state.c | 23 +++++++++++++------
 drivers/gpu/drm/i915/i915_gem_render_state.h | 18 ---------------
 drivers/gpu/drm/i915/intel_lrc.c             | 34 +---------------------------
 drivers/gpu/drm/i915/intel_renderstate.h     | 16 +++++++++----
 4 files changed, 28 insertions(+), 63 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index 2ba759f3ab6f..a9b56d18a93b 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -28,6 +28,15 @@
 #include "i915_drv.h"
 #include "intel_renderstate.h"
 
+struct render_state {
+	const struct intel_renderstate_rodata *rodata;
+	struct drm_i915_gem_object *obj;
+	u64 ggtt_offset;
+	int gen;
+	u32 aux_batch_size;
+	u32 aux_batch_offset;
+};
+
 static const struct intel_renderstate_rodata *
 render_state_get_rodata(const int gen)
 {
@@ -51,6 +60,7 @@ static int render_state_init(struct render_state *so,
 	int ret;
 
 	so->gen = INTEL_GEN(dev_priv);
+	so->ggtt_offset = 0; /* keep gcc quiet */
 	so->rodata = render_state_get_rodata(so->gen);
 	if (so->rodata == NULL)
 		return 0;
@@ -192,14 +202,14 @@ err_out:
 
 #undef OUT_BATCH
 
-void i915_gem_render_state_fini(struct render_state *so)
+static void render_state_fini(struct render_state *so)
 {
 	i915_gem_object_ggtt_unpin(so->obj);
 	i915_gem_object_put(so->obj);
 }
 
-int i915_gem_render_state_prepare(struct intel_engine_cs *engine,
-				  struct render_state *so)
+static int render_state_prepare(struct intel_engine_cs *engine,
+				struct render_state *so)
 {
 	int ret;
 
@@ -215,7 +225,7 @@ int i915_gem_render_state_prepare(struct intel_engine_cs *engine,
 
 	ret = render_state_setup(so);
 	if (ret) {
-		i915_gem_render_state_fini(so);
+		render_state_fini(so);
 		return ret;
 	}
 
@@ -227,7 +237,7 @@ int i915_gem_render_state_init(struct drm_i915_gem_request *req)
 	struct render_state so;
 	int ret;
 
-	ret = i915_gem_render_state_prepare(req->engine, &so);
+	ret = render_state_prepare(req->engine, &so);
 	if (ret)
 		return ret;
 
@@ -251,8 +261,7 @@ int i915_gem_render_state_init(struct drm_i915_gem_request *req)
 	}
 
 	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), req);
-
 out:
-	i915_gem_render_state_fini(&so);
+	render_state_fini(&so);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.h b/drivers/gpu/drm/i915/i915_gem_render_state.h
index 6aaa3a10a630..c44fca8599bb 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.h
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.h
@@ -26,24 +26,6 @@
 
 #include <linux/types.h>
 
-struct intel_renderstate_rodata {
-	const u32 *reloc;
-	const u32 *batch;
-	const u32 batch_items;
-};
-
-struct render_state {
-	const struct intel_renderstate_rodata *rodata;
-	struct drm_i915_gem_object *obj;
-	u64 ggtt_offset;
-	int gen;
-	u32 aux_batch_size;
-	u32 aux_batch_offset;
-};
-
 int i915_gem_render_state_init(struct drm_i915_gem_request *req);
-void i915_gem_render_state_fini(struct render_state *so);
-int i915_gem_render_state_prepare(struct intel_engine_cs *engine,
-				  struct render_state *so);
 
 #endif /* _I915_GEM_RENDER_STATE_H_ */
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 149a0dc7aeed..7d54dac3ce22 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1803,38 +1803,6 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request)
 	return intel_logical_ring_advance(request);
 }
 
-static int intel_lr_context_render_state_init(struct drm_i915_gem_request *req)
-{
-	struct render_state so;
-	int ret;
-
-	ret = i915_gem_render_state_prepare(req->engine, &so);
-	if (ret)
-		return ret;
-
-	if (so.rodata == NULL)
-		return 0;
-
-	ret = req->engine->emit_bb_start(req, so.ggtt_offset,
-					 so.rodata->batch_items * 4,
-					 I915_DISPATCH_SECURE);
-	if (ret)
-		goto out;
-
-	ret = req->engine->emit_bb_start(req,
-					 (so.ggtt_offset + so.aux_batch_offset),
-					 so.aux_batch_size,
-					 I915_DISPATCH_SECURE);
-	if (ret)
-		goto out;
-
-	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), req);
-
-out:
-	i915_gem_render_state_fini(&so);
-	return ret;
-}
-
 static int gen8_init_rcs_context(struct drm_i915_gem_request *req)
 {
 	int ret;
@@ -1851,7 +1819,7 @@ static int gen8_init_rcs_context(struct drm_i915_gem_request *req)
 	if (ret)
 		DRM_ERROR("MOCS failed to program: expect performance issues.\n");
 
-	return intel_lr_context_render_state_init(req);
+	return i915_gem_render_state_init(req);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/intel_renderstate.h b/drivers/gpu/drm/i915/intel_renderstate.h
index 5bd69852752c..08f6fea05a2c 100644
--- a/drivers/gpu/drm/i915/intel_renderstate.h
+++ b/drivers/gpu/drm/i915/intel_renderstate.h
@@ -24,12 +24,13 @@
 #ifndef _INTEL_RENDERSTATE_H
 #define _INTEL_RENDERSTATE_H
 
-#include "i915_drv.h"
+#include <linux/types.h>
 
-extern const struct intel_renderstate_rodata gen6_null_state;
-extern const struct intel_renderstate_rodata gen7_null_state;
-extern const struct intel_renderstate_rodata gen8_null_state;
-extern const struct intel_renderstate_rodata gen9_null_state;
+struct intel_renderstate_rodata {
+	const u32 *reloc;
+	const u32 *batch;
+	const u32 batch_items;
+};
 
 #define RO_RENDERSTATE(_g)						\
 	const struct intel_renderstate_rodata gen ## _g ## _null_state = { \
@@ -38,4 +39,9 @@ extern const struct intel_renderstate_rodata gen9_null_state;
 		.batch_items = sizeof(gen ## _g ## _null_state_batch)/4, \
 	}
 
+extern const struct intel_renderstate_rodata gen6_null_state;
+extern const struct intel_renderstate_rodata gen7_null_state;
+extern const struct intel_renderstate_rodata gen8_null_state;
+extern const struct intel_renderstate_rodata gen9_null_state;
+
 #endif /* INTEL_RENDERSTATE_H */
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 28/31] drm/i915: Refactor golden render state emission to unconfuse gcc

[Chris Wilson]

GCC was inlining the init and setup functions, but was getting itself
confused into thinking that variables could be used uninitialised. If we
do the inline for gcc, it is happy! As a bonus we shrink the code.

v2: A couple of minor tweaks from Joonas

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_render_state.c | 95 ++++++++--------------------
 1 file changed, 27 insertions(+), 68 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index a9b56d18a93b..f85c5505bce2 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -32,15 +32,14 @@ struct render_state {
 	const struct intel_renderstate_rodata *rodata;
 	struct drm_i915_gem_object *obj;
 	u64 ggtt_offset;
-	int gen;
 	u32 aux_batch_size;
 	u32 aux_batch_offset;
 };
 
 static const struct intel_renderstate_rodata *
-render_state_get_rodata(const int gen)
+render_state_get_rodata(const struct drm_i915_gem_request *req)
 {
-	switch (gen) {
+	switch (INTEL_GEN(req->i915)) {
 	case 6:
 		return &gen6_null_state;
 	case 7:
@@ -54,36 +53,6 @@ render_state_get_rodata(const int gen)
 	return NULL;
 }
 
-static int render_state_init(struct render_state *so,
-			     struct drm_i915_private *dev_priv)
-{
-	int ret;
-
-	so->gen = INTEL_GEN(dev_priv);
-	so->ggtt_offset = 0; /* keep gcc quiet */
-	so->rodata = render_state_get_rodata(so->gen);
-	if (so->rodata == NULL)
-		return 0;
-
-	if (so->rodata->batch_items * 4 > 4096)
-		return -EINVAL;
-
-	so->obj = i915_gem_object_create(&dev_priv->drm, 4096);
-	if (IS_ERR(so->obj))
-		return PTR_ERR(so->obj);
-
-	ret = i915_gem_obj_ggtt_pin(so->obj, 4096, 0);
-	if (ret)
-		goto free_gem;
-
-	so->ggtt_offset = i915_gem_obj_ggtt_offset(so->obj);
-	return 0;
-
-free_gem:
-	i915_gem_object_put(so->obj);
-	return ret;
-}
-
 /*
  * Macro to add commands to auxiliary batch.
  * This macro only checks for page overflow before inserting the commands,
@@ -106,6 +75,7 @@ static int render_state_setup(struct render_state *so)
 {
 	struct drm_device *dev = so->obj->base.dev;
 	const struct intel_renderstate_rodata *rodata = so->rodata;
+	const bool has_64bit_reloc = INTEL_GEN(dev) >= 8;
 	unsigned int i = 0, reloc_index = 0;
 	struct page *page;
 	u32 *d;
@@ -124,7 +94,7 @@ static int render_state_setup(struct render_state *so)
 		if (i * 4  == rodata->reloc[reloc_index]) {
 			u64 r = s + so->ggtt_offset;
 			s = lower_32_bits(r);
-			if (so->gen >= 8) {
+			if (has_64bit_reloc) {
 				if (i + 1 >= rodata->batch_items ||
 				    rodata->batch[i + 1] != 0) {
 					ret = -EINVAL;
@@ -202,53 +172,40 @@ err_out:
 
 #undef OUT_BATCH
 
-static void render_state_fini(struct render_state *so)
-{
-	i915_gem_object_ggtt_unpin(so->obj);
-	i915_gem_object_put(so->obj);
-}
-
-static int render_state_prepare(struct intel_engine_cs *engine,
-				struct render_state *so)
+int i915_gem_render_state_init(struct drm_i915_gem_request *req)
 {
+	struct render_state so;
 	int ret;
 
-	if (WARN_ON(engine->id != RCS))
+	if (WARN_ON(req->engine->id != RCS))
 		return -ENOENT;
 
-	ret = render_state_init(so, engine->i915);
-	if (ret)
-		return ret;
-
-	if (so->rodata == NULL)
+	so.rodata = render_state_get_rodata(req);
+	if (!so.rodata)
 		return 0;
 
-	ret = render_state_setup(so);
-	if (ret) {
-		render_state_fini(so);
-		return ret;
-	}
-
-	return 0;
-}
+	if (so.rodata->batch_items * 4 > 4096)
+		return -EINVAL;
 
-int i915_gem_render_state_init(struct drm_i915_gem_request *req)
-{
-	struct render_state so;
-	int ret;
+	so.obj = i915_gem_object_create(&req->i915->drm, 4096);
+	if (IS_ERR(so.obj))
+		return PTR_ERR(so.obj);
 
-	ret = render_state_prepare(req->engine, &so);
+	ret = i915_gem_obj_ggtt_pin(so.obj, 4096, 0);
 	if (ret)
-		return ret;
+		goto err_obj;
 
-	if (so.rodata == NULL)
-		return 0;
+	so.ggtt_offset = i915_gem_obj_ggtt_offset(so.obj);
+
+	ret = render_state_setup(&so);
+	if (ret)
+		goto err_unpin;
 
 	ret = req->engine->emit_bb_start(req, so.ggtt_offset,
 					 so.rodata->batch_items * 4,
 					 I915_DISPATCH_SECURE);
 	if (ret)
-		goto out;
+		goto err_unpin;
 
 	if (so.aux_batch_size > 8) {
 		ret = req->engine->emit_bb_start(req,
@@ -257,11 +214,13 @@ int i915_gem_render_state_init(struct drm_i915_gem_request *req)
 						 so.aux_batch_size,
 						 I915_DISPATCH_SECURE);
 		if (ret)
-			goto out;
+			goto err_unpin;
 	}
 
 	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), req);
-out:
-	render_state_fini(&so);
+err_unpin:
+	i915_gem_object_ggtt_unpin(so.obj);
+err_obj:
+	i915_gem_object_put(so.obj);
 	return ret;
 }
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 29/31] drm/i915: Unify legacy/execlists submit_execbuf callbacks

[Chris Wilson]

Now that emitting requests is identical between legacy and execlists, we
can use the same function to build up the ring for submitting to either
engine. (With the exception of i915_switch_contexts(), but in time that
will also be handled gracefully.)

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h            |  20 -----
 drivers/gpu/drm/i915/i915_gem.c            |   2 -
 drivers/gpu/drm/i915/i915_gem_context.c    |   7 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  24 ++++--
 drivers/gpu/drm/i915/intel_lrc.c           | 123 -----------------------------
 drivers/gpu/drm/i915/intel_lrc.h           |   4 -
 drivers/gpu/drm/i915/intel_ringbuffer.c    |   6 +-
 7 files changed, 23 insertions(+), 163 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 1031f5a7126a..3c0c31398e7e 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1705,18 +1705,6 @@ struct i915_virtual_gpu {
 	bool active;
 };
 
-struct i915_execbuffer_params {
-	struct drm_device               *dev;
-	struct drm_file                 *file;
-	uint32_t                        dispatch_flags;
-	uint32_t                        args_batch_start_offset;
-	uint64_t                        batch_obj_vm_offset;
-	struct intel_engine_cs *engine;
-	struct drm_i915_gem_object      *batch_obj;
-	struct i915_gem_context            *ctx;
-	struct drm_i915_gem_request     *request;
-};
-
 /* used in computing the new watermarks state */
 struct intel_wm_config {
 	unsigned int num_pipes_active;
@@ -2016,9 +2004,6 @@ struct drm_i915_private {
 
 	/* Abstract the submission mechanism (legacy ringbuffer or execlists) away */
 	struct {
-		int (*execbuf_submit)(struct i915_execbuffer_params *params,
-				      struct drm_i915_gem_execbuffer2 *args,
-				      struct list_head *vmas);
 		void (*cleanup_engine)(struct intel_engine_cs *engine);
 		void (*stop_engine)(struct intel_engine_cs *engine);
 
@@ -2993,11 +2978,6 @@ int i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
 int i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
-void i915_gem_execbuffer_move_to_active(struct list_head *vmas,
-					struct drm_i915_gem_request *req);
-int i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
-				   struct drm_i915_gem_execbuffer2 *args,
-				   struct list_head *vmas);
 int i915_gem_execbuffer(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 int i915_gem_execbuffer2(struct drm_device *dev, void *data,
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index b6c4ff63725f..d79b949fb4c4 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -4534,11 +4534,9 @@ int i915_gem_init(struct drm_device *dev)
 	mutex_lock(&dev->struct_mutex);
 
 	if (!i915.enable_execlists) {
-		dev_priv->gt.execbuf_submit = i915_gem_ringbuffer_submission;
 		dev_priv->gt.cleanup_engine = intel_engine_cleanup;
 		dev_priv->gt.stop_engine = intel_engine_stop;
 	} else {
-		dev_priv->gt.execbuf_submit = intel_execlists_submission;
 		dev_priv->gt.cleanup_engine = intel_logical_ring_cleanup;
 		dev_priv->gt.stop_engine = intel_logical_ring_stop;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index beece8feb8fe..4e233dffb592 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -894,8 +894,9 @@ int i915_switch_context(struct drm_i915_gem_request *req)
 {
 	struct intel_engine_cs *engine = req->engine;
 
-	WARN_ON(i915.enable_execlists);
 	lockdep_assert_held(&req->i915->drm.struct_mutex);
+	if (i915.enable_execlists)
+		return 0;
 
 	if (!req->ctx->engine[engine->id].state) {
 		struct i915_gem_context *to = req->ctx;
@@ -943,9 +944,7 @@ int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 		if (IS_ERR(req))
 			return PTR_ERR(req);
 
-		ret = 0;
-		if (!i915.enable_execlists)
-			ret = i915_switch_context(req);
+		ret = i915_switch_context(req);
 		i915_add_request_no_flush(req);
 		if (ret)
 			return ret;
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 1c0e6f1075b7..af61b7b00681 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -42,6 +42,18 @@
 
 #define BATCH_OFFSET_BIAS (256*1024)
 
+struct i915_execbuffer_params {
+	struct drm_device               *dev;
+	struct drm_file                 *file;
+	u32				 dispatch_flags;
+	u32				 args_batch_start_offset;
+	u32				 batch_obj_vm_offset;
+	struct intel_engine_cs          *engine;
+	struct drm_i915_gem_object      *batch_obj;
+	struct i915_gem_context         *ctx;
+	struct drm_i915_gem_request     *request;
+};
+
 struct eb_vmas {
 	struct list_head vmas;
 	int and;
@@ -1117,7 +1129,7 @@ i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
 	return ctx;
 }
 
-void
+static void
 i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 				   struct drm_i915_gem_request *req)
 {
@@ -1244,10 +1256,10 @@ err:
 		return ERR_PTR(ret);
 }
 
-int
-i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
-			       struct drm_i915_gem_execbuffer2 *args,
-			       struct list_head *vmas)
+static int
+execbuf_submit(struct i915_execbuffer_params *params,
+	       struct drm_i915_gem_execbuffer2 *args,
+	       struct list_head *vmas)
 {
 	struct drm_i915_private *dev_priv = params->request->i915;
 	u64 exec_start, exec_len;
@@ -1637,7 +1649,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	params->batch_obj               = batch_obj;
 	params->ctx                     = ctx;
 
-	ret = dev_priv->gt.execbuf_submit(params, args, &eb->vmas);
+	ret = execbuf_submit(params, args, &eb->vmas);
 err_request:
 	i915_gem_execbuffer_retire_commands(params);
 
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 7d54dac3ce22..c4b507ba9369 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -642,39 +642,6 @@ static void execlists_submit_request(struct drm_i915_gem_request *request)
 	spin_unlock_bh(&engine->execlist_lock);
 }
 
-static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
-				 struct list_head *vmas)
-{
-	const unsigned other_rings = ~intel_engine_flag(req->engine);
-	struct i915_vma *vma;
-	uint32_t flush_domains = 0;
-	bool flush_chipset = false;
-	int ret;
-
-	list_for_each_entry(vma, vmas, exec_list) {
-		struct drm_i915_gem_object *obj = vma->obj;
-
-		if (obj->active & other_rings) {
-			ret = i915_gem_object_sync(obj, req);
-			if (ret)
-				return ret;
-		}
-
-		if (obj->base.write_domain & I915_GEM_DOMAIN_CPU)
-			flush_chipset |= i915_gem_clflush_object(obj, false);
-
-		flush_domains |= obj->base.write_domain;
-	}
-
-	if (flush_domains & I915_GEM_DOMAIN_GTT)
-		wmb();
-
-	/* Unconditionally invalidate gpu caches and ensure that we do flush
-	 * any residual writes from the previous batch.
-	 */
-	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
-}
-
 int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
@@ -776,96 +743,6 @@ intel_logical_ring_advance(struct drm_i915_gem_request *request)
 	return 0;
 }
 
-/**
- * intel_execlists_submission() - submit a batchbuffer for execution, Execlists style
- * @params: execbuffer call parameters.
- * @args: execbuffer call arguments.
- * @vmas: list of vmas.
- *
- * This is the evil twin version of i915_gem_ringbuffer_submission. It abstracts
- * away the submission details of the execbuffer ioctl call.
- *
- * Return: non-zero if the submission fails.
- */
-int intel_execlists_submission(struct i915_execbuffer_params *params,
-			       struct drm_i915_gem_execbuffer2 *args,
-			       struct list_head *vmas)
-{
-	struct drm_device       *dev = params->dev;
-	struct intel_engine_cs *engine = params->engine;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_ring *ring = params->request->ring;
-	u64 exec_start;
-	int instp_mode;
-	u32 instp_mask;
-	int ret;
-
-	instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK;
-	instp_mask = I915_EXEC_CONSTANTS_MASK;
-	switch (instp_mode) {
-	case I915_EXEC_CONSTANTS_REL_GENERAL:
-	case I915_EXEC_CONSTANTS_ABSOLUTE:
-	case I915_EXEC_CONSTANTS_REL_SURFACE:
-		if (instp_mode != 0 && engine->id != RCS) {
-			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
-			return -EINVAL;
-		}
-
-		if (instp_mode != dev_priv->relative_constants_mode) {
-			if (instp_mode == I915_EXEC_CONSTANTS_REL_SURFACE) {
-				DRM_DEBUG("rel surface constants mode invalid on gen5+\n");
-				return -EINVAL;
-			}
-
-			/* The HW changed the meaning on this bit on gen6 */
-			instp_mask &= ~I915_EXEC_CONSTANTS_REL_SURFACE;
-		}
-		break;
-	default:
-		DRM_DEBUG("execbuf with unknown constants: %d\n", instp_mode);
-		return -EINVAL;
-	}
-
-	if (args->flags & I915_EXEC_GEN7_SOL_RESET) {
-		DRM_DEBUG("sol reset is gen7 only\n");
-		return -EINVAL;
-	}
-
-	ret = execlists_move_to_gpu(params->request, vmas);
-	if (ret)
-		return ret;
-
-	if (engine->id == RCS &&
-	    instp_mode != dev_priv->relative_constants_mode) {
-		ret = intel_ring_begin(params->request, 4);
-		if (ret)
-			return ret;
-
-		intel_ring_emit(ring, MI_NOOP);
-		intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
-		intel_ring_emit_reg(ring, INSTPM);
-		intel_ring_emit(ring, instp_mask << 16 | instp_mode);
-		intel_ring_advance(ring);
-
-		dev_priv->relative_constants_mode = instp_mode;
-	}
-
-	exec_start = params->batch_obj_vm_offset +
-		     args->batch_start_offset;
-
-	ret = engine->emit_bb_start(params->request,
-				    exec_start, args->batch_len,
-				    params->dispatch_flags);
-	if (ret)
-		return ret;
-
-	trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags);
-
-	i915_gem_execbuffer_move_to_active(vmas, params->request);
-
-	return 0;
-}
-
 void intel_execlists_cancel_requests(struct intel_engine_cs *engine)
 {
 	struct drm_i915_gem_request *req, *tmp;
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index 33e0193e5451..a1d4e7f9c64c 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -95,10 +95,6 @@ uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx,
 /* Execlists */
 int intel_sanitize_enable_execlists(struct drm_i915_private *dev_priv,
 				    int enable_execlists);
-struct i915_execbuffer_params;
-int intel_execlists_submission(struct i915_execbuffer_params *params,
-			       struct drm_i915_gem_execbuffer2 *args,
-			       struct list_head *vmas);
 
 void intel_execlists_cancel_requests(struct intel_engine_cs *engine);
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 0fe4b000749a..189f29e06dfd 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -2493,9 +2493,7 @@ static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
 		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
 
 	/* Now that the ring is fully powered up, update the tail */
-	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
-		      request->tail % (request->ring->size - 1));
-	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
+	i9xx_submit_request(request);
 
 	/* Let the ring send IDLE messages to the GT again,
 	 * and so let it sleep to conserve power when idle.
@@ -2806,9 +2804,9 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 	engine->init_hw = init_ring_common;
 
 	engine->emit_request = i9xx_emit_request;
-	engine->submit_request = i9xx_submit_request;
 	if (i915.semaphores)
 		engine->emit_request = gen6_sema_emit_request;
+	engine->submit_request = i9xx_submit_request;
 
 	if (INTEL_GEN(dev_priv) >= 8)
 		engine->emit_bb_start = gen8_emit_bb_start;
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 30/31] drm/i915: Simplify calling engine->sync_to

[Chris Wilson]

Since requests can no longer be generated as a side-effect of
intel_ring_begin(), we know that the seqno will be unchanged during
ring-emission. This predicatablity then means we do not have to check
for the seqno wrapping around whilst emitting the semaphore for
engine->sync_to().

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h         |  2 +-
 drivers/gpu/drm/i915/i915_gem.c         | 13 ++-----
 drivers/gpu/drm/i915/i915_gem_request.c |  9 +----
 drivers/gpu/drm/i915/intel_ringbuffer.c | 64 ++++++++++++---------------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  5 ++-
 5 files changed, 30 insertions(+), 63 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 3c0c31398e7e..f91f72bc6e8d 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1757,7 +1757,7 @@ struct drm_i915_private {
 	struct i915_gem_context *kernel_context;
 	struct intel_engine_cs engine[I915_NUM_ENGINES];
 	struct drm_i915_gem_object *semaphore_obj;
-	uint32_t last_seqno, next_seqno;
+	u32 next_seqno;
 
 	struct drm_dma_handle *status_page_dmah;
 	struct resource mch_res;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index d79b949fb4c4..3df6b485d2d4 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2867,22 +2867,15 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
 		i915_gem_object_retire_request(obj, from);
 	} else {
 		int idx = intel_engine_sync_index(from->engine, to->engine);
-		u32 seqno = i915_gem_request_get_seqno(from);
-
-		if (seqno <= from->engine->semaphore.sync_seqno[idx])
+		if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx])
 			return 0;
 
 		trace_i915_gem_ring_sync_to(to, from);
-		ret = to->engine->semaphore.sync_to(to, from->engine, seqno);
+		ret = to->engine->semaphore.sync_to(to, from);
 		if (ret)
 			return ret;
 
-		/* We use last_read_req because sync_to()
-		 * might have just caused seqno wrap under
-		 * the radar.
-		 */
-		from->engine->semaphore.sync_seqno[idx] =
-			i915_gem_request_get_seqno(obj->last_read_req[from->engine->id]);
+		from->engine->semaphore.sync_seqno[idx] = from->fence.seqno;
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index f4bf9f669eed..aaced4892684 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -264,14 +264,7 @@ int i915_gem_set_seqno(struct drm_device *dev, u32 seqno)
 	if (ret)
 		return ret;
 
-	/* Carefully set the last_seqno value so that wrap
-	 * detection still works
-	 */
 	dev_priv->next_seqno = seqno;
-	dev_priv->last_seqno = seqno - 1;
-	if (dev_priv->last_seqno == 0)
-		dev_priv->last_seqno--;
-
 	return 0;
 }
 
@@ -288,7 +281,7 @@ static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno)
 		dev_priv->next_seqno = 1;
 	}
 
-	*seqno = dev_priv->last_seqno = dev_priv->next_seqno++;
+	*seqno = dev_priv->next_seqno++;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 189f29e06dfd..e35088c8a3ef 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1509,12 +1509,6 @@ static int gen8_render_emit_request(struct drm_i915_gem_request *req)
 	return 0;
 }
 
-static inline bool i915_gem_has_seqno_wrapped(struct drm_i915_private *dev_priv,
-					      u32 seqno)
-{
-	return dev_priv->last_seqno < seqno;
-}
-
 /**
  * intel_ring_sync - sync the waiter to the signaller on seqno
  *
@@ -1524,24 +1518,23 @@ static inline bool i915_gem_has_seqno_wrapped(struct drm_i915_private *dev_priv,
  */
 
 static int
-gen8_ring_sync(struct drm_i915_gem_request *waiter_req,
-	       struct intel_engine_cs *signaller,
-	       u32 seqno)
+gen8_ring_sync(struct drm_i915_gem_request *wait,
+	       struct drm_i915_gem_request *signal)
 {
-	struct intel_ring *waiter = waiter_req->ring;
-	struct drm_i915_private *dev_priv = waiter_req->i915;
-	u64 offset = GEN8_WAIT_OFFSET(waiter_req->engine, signaller->id);
+	struct intel_ring *waiter = wait->ring;
+	struct drm_i915_private *dev_priv = wait->i915;
+	u64 offset = GEN8_WAIT_OFFSET(wait->engine, signal->engine->id);
 	struct i915_hw_ppgtt *ppgtt;
 	int ret;
 
-	ret = intel_ring_begin(waiter_req, 4);
+	ret = intel_ring_begin(wait, 4);
 	if (ret)
 		return ret;
 
 	intel_ring_emit(waiter, MI_SEMAPHORE_WAIT |
 				MI_SEMAPHORE_GLOBAL_GTT |
 				MI_SEMAPHORE_SAD_GTE_SDD);
-	intel_ring_emit(waiter, seqno);
+	intel_ring_emit(waiter, signal->fence.seqno);
 	intel_ring_emit(waiter, lower_32_bits(offset));
 	intel_ring_emit(waiter, upper_32_bits(offset));
 	intel_ring_advance(waiter);
@@ -1551,48 +1544,37 @@ gen8_ring_sync(struct drm_i915_gem_request *waiter_req,
 	 * We do this on the i915_switch_context() following the wait and
 	 * before the dispatch.
 	 */
-	ppgtt = waiter_req->ctx->ppgtt;
-	if (ppgtt && waiter_req->engine->id != RCS)
-		ppgtt->pd_dirty_rings |= intel_engine_flag(waiter_req->engine);
+	ppgtt = wait->ctx->ppgtt;
+	if (ppgtt && wait->engine->id != RCS)
+		ppgtt->pd_dirty_rings |= intel_engine_flag(wait->engine);
 	return 0;
 }
 
 static int
-gen6_ring_sync(struct drm_i915_gem_request *waiter_req,
-	       struct intel_engine_cs *signaller,
-	       u32 seqno)
+gen6_ring_sync(struct drm_i915_gem_request *wait,
+	       struct drm_i915_gem_request *signal)
 {
-	struct intel_ring *waiter = waiter_req->ring;
+	struct intel_ring *waiter = wait->ring;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
-	u32 wait_mbox = signaller->semaphore.mbox.wait[waiter_req->engine->id];
+	u32 wait_mbox = signal->engine->semaphore.mbox.wait[wait->engine->id];
 	int ret;
 
-	/* Throughout all of the GEM code, seqno passed implies our current
-	 * seqno is >= the last seqno executed. However for hardware the
-	 * comparison is strictly greater than.
-	 */
-	seqno -= 1;
-
 	WARN_ON(wait_mbox == MI_SEMAPHORE_SYNC_INVALID);
 
-	ret = intel_ring_begin(waiter_req, 4);
+	ret = intel_ring_begin(wait, 4);
 	if (ret)
 		return ret;
 
-	/* If seqno wrap happened, omit the wait with no-ops */
-	if (likely(!i915_gem_has_seqno_wrapped(waiter_req->i915, seqno))) {
-		intel_ring_emit(waiter, dw1 | wait_mbox);
-		intel_ring_emit(waiter, seqno);
-		intel_ring_emit(waiter, 0);
-		intel_ring_emit(waiter, MI_NOOP);
-	} else {
-		intel_ring_emit(waiter, MI_NOOP);
-		intel_ring_emit(waiter, MI_NOOP);
-		intel_ring_emit(waiter, MI_NOOP);
-		intel_ring_emit(waiter, MI_NOOP);
-	}
+	intel_ring_emit(waiter, dw1 | wait_mbox);
+	/* Throughout all of the GEM code, seqno passed implies our current
+	 * seqno is >= the last seqno executed. However for hardware the
+	 * comparison is strictly greater than.
+	 */
+	intel_ring_emit(waiter, signal->fence.seqno - 1);
+	intel_ring_emit(waiter, 0);
+	intel_ring_emit(waiter, MI_NOOP);
 	intel_ring_advance(waiter);
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index a137fd218c83..7a4cf4a14c1a 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -275,9 +275,8 @@ struct intel_engine_cs {
 		};
 
 		/* AKA wait() */
-		int	(*sync_to)(struct drm_i915_gem_request *to_req,
-				   struct intel_engine_cs *from,
-				   u32 seqno);
+		int	(*sync_to)(struct drm_i915_gem_request *to,
+				   struct drm_i915_gem_request *from);
 		int	(*signal)(struct drm_i915_gem_request *signaller_req);
 	} semaphore;
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 31/31] drm/i915: Rename engine->semaphore.sync_to, engine->sempahore.signal locals

[Chris Wilson]

In order to be more consistent with the rest of the request construction
and ring emission, use the common names for the ring and request.

Rather than using signaler_req, waiter_req, and intel_ring *wait, we use
plain req and ring.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 132 ++++++++++++++++----------------
 drivers/gpu/drm/i915/intel_ringbuffer.h |   6 +-
 2 files changed, 68 insertions(+), 70 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index e35088c8a3ef..1d94779ab2fd 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1322,108 +1322,105 @@ static void render_ring_cleanup(struct intel_engine_cs *engine)
 	intel_fini_pipe_control(engine);
 }
 
-static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req)
+static int gen8_rcs_signal(struct drm_i915_gem_request *req)
 {
-	struct intel_ring *signaller = signaller_req->ring;
-	struct drm_i915_private *dev_priv = signaller_req->i915;
+	struct intel_ring *ring = req->ring;
+	struct drm_i915_private *dev_priv = req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
 	int ret, num_rings;
 
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	ret = intel_ring_begin(signaller_req, (num_rings-1) * 8);
+	ret = intel_ring_begin(req, (num_rings-1) * 8);
 	if (ret)
 		return ret;
 
 	for_each_engine_id(waiter, dev_priv, id) {
-		u64 gtt_offset =
-			signaller_req->engine->semaphore.signal_ggtt[id];
+		u64 gtt_offset = req->engine->semaphore.signal_ggtt[id];
 		if (gtt_offset == MI_SEMAPHORE_SYNC_INVALID)
 			continue;
 
-		intel_ring_emit(signaller, GFX_OP_PIPE_CONTROL(6));
-		intel_ring_emit(signaller,
+		intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(6));
+		intel_ring_emit(ring,
 				PIPE_CONTROL_GLOBAL_GTT_IVB |
 				PIPE_CONTROL_QW_WRITE |
 				PIPE_CONTROL_CS_STALL);
-		intel_ring_emit(signaller, lower_32_bits(gtt_offset));
-		intel_ring_emit(signaller, upper_32_bits(gtt_offset));
-		intel_ring_emit(signaller, signaller_req->fence.seqno);
-		intel_ring_emit(signaller, 0);
-		intel_ring_emit(signaller,
+		intel_ring_emit(ring, lower_32_bits(gtt_offset));
+		intel_ring_emit(ring, upper_32_bits(gtt_offset));
+		intel_ring_emit(ring, req->fence.seqno);
+		intel_ring_emit(ring, 0);
+		intel_ring_emit(ring,
 				MI_SEMAPHORE_SIGNAL |
 				MI_SEMAPHORE_TARGET(waiter->hw_id));
-		intel_ring_emit(signaller, 0);
+		intel_ring_emit(ring, 0);
 	}
-	intel_ring_advance(signaller);
+	intel_ring_advance(ring);
 
 	return 0;
 }
 
-static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req)
+static int gen8_xcs_signal(struct drm_i915_gem_request *req)
 {
-	struct intel_ring *signaller = signaller_req->ring;
-	struct drm_i915_private *dev_priv = signaller_req->i915;
+	struct intel_ring *ring = req->ring;
+	struct drm_i915_private *dev_priv = req->i915;
 	struct intel_engine_cs *waiter;
 	enum intel_engine_id id;
 	int ret, num_rings;
 
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	ret = intel_ring_begin(signaller_req, (num_rings-1) * 6);
+	ret = intel_ring_begin(req, (num_rings-1) * 6);
 	if (ret)
 		return ret;
 
 	for_each_engine_id(waiter, dev_priv, id) {
-		u64 gtt_offset =
-			signaller_req->engine->semaphore.signal_ggtt[id];
+		u64 gtt_offset = req->engine->semaphore.signal_ggtt[id];
 		if (gtt_offset == MI_SEMAPHORE_SYNC_INVALID)
 			continue;
 
-		intel_ring_emit(signaller,
+		intel_ring_emit(ring,
 				(MI_FLUSH_DW + 1) | MI_FLUSH_DW_OP_STOREDW);
-		intel_ring_emit(signaller,
+		intel_ring_emit(ring,
 				lower_32_bits(gtt_offset) |
 				MI_FLUSH_DW_USE_GTT);
-		intel_ring_emit(signaller, upper_32_bits(gtt_offset));
-		intel_ring_emit(signaller, signaller_req->fence.seqno);
-		intel_ring_emit(signaller,
+		intel_ring_emit(ring, upper_32_bits(gtt_offset));
+		intel_ring_emit(ring, req->fence.seqno);
+		intel_ring_emit(ring,
 				MI_SEMAPHORE_SIGNAL |
 				MI_SEMAPHORE_TARGET(waiter->hw_id));
-		intel_ring_emit(signaller, 0);
+		intel_ring_emit(ring, 0);
 	}
-	intel_ring_advance(signaller);
+	intel_ring_advance(ring);
 
 	return 0;
 }
 
-static int gen6_signal(struct drm_i915_gem_request *signaller_req)
+static int gen6_signal(struct drm_i915_gem_request *req)
 {
-	struct intel_ring *signaller = signaller_req->ring;
-	struct drm_i915_private *dev_priv = signaller_req->i915;
+	struct intel_ring *ring = req->ring;
+	struct drm_i915_private *dev_priv = req->i915;
 	struct intel_engine_cs *useless;
 	enum intel_engine_id id;
 	int ret, num_rings;
 
 	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
-	ret = intel_ring_begin(signaller_req, round_up((num_rings-1) * 3, 2));
+	ret = intel_ring_begin(req, round_up((num_rings-1) * 3, 2));
 	if (ret)
 		return ret;
 
 	for_each_engine_id(useless, dev_priv, id) {
-		i915_reg_t mbox_reg =
-			signaller_req->engine->semaphore.mbox.signal[id];
+		i915_reg_t mbox_reg = req->engine->semaphore.mbox.signal[id];
 
 		if (i915_mmio_reg_valid(mbox_reg)) {
-			intel_ring_emit(signaller, MI_LOAD_REGISTER_IMM(1));
-			intel_ring_emit_reg(signaller, mbox_reg);
-			intel_ring_emit(signaller, signaller_req->fence.seqno);
+			intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
+			intel_ring_emit_reg(ring, mbox_reg);
+			intel_ring_emit(ring, req->fence.seqno);
 		}
 	}
 
 	/* If num_dwords was rounded, make sure the tail pointer is correct */
 	if (num_rings % 2 == 0)
-		intel_ring_emit(signaller, MI_NOOP);
-	intel_ring_advance(signaller);
+		intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -1518,64 +1515,65 @@ static int gen8_render_emit_request(struct drm_i915_gem_request *req)
  */
 
 static int
-gen8_ring_sync(struct drm_i915_gem_request *wait,
-	       struct drm_i915_gem_request *signal)
+gen8_ring_sync_to(struct drm_i915_gem_request *req,
+		  struct drm_i915_gem_request *signal)
 {
-	struct intel_ring *waiter = wait->ring;
-	struct drm_i915_private *dev_priv = wait->i915;
-	u64 offset = GEN8_WAIT_OFFSET(wait->engine, signal->engine->id);
+	struct intel_ring *ring = req->ring;
+	struct drm_i915_private *dev_priv = req->i915;
+	u64 offset = GEN8_WAIT_OFFSET(req->engine, signal->engine->id);
 	struct i915_hw_ppgtt *ppgtt;
 	int ret;
 
-	ret = intel_ring_begin(wait, 4);
+	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(waiter, MI_SEMAPHORE_WAIT |
-				MI_SEMAPHORE_GLOBAL_GTT |
-				MI_SEMAPHORE_SAD_GTE_SDD);
-	intel_ring_emit(waiter, signal->fence.seqno);
-	intel_ring_emit(waiter, lower_32_bits(offset));
-	intel_ring_emit(waiter, upper_32_bits(offset));
-	intel_ring_advance(waiter);
+	intel_ring_emit(ring,
+			MI_SEMAPHORE_WAIT |
+			MI_SEMAPHORE_GLOBAL_GTT |
+			MI_SEMAPHORE_SAD_GTE_SDD);
+	intel_ring_emit(ring, signal->fence.seqno);
+	intel_ring_emit(ring, lower_32_bits(offset));
+	intel_ring_emit(ring, upper_32_bits(offset));
+	intel_ring_advance(ring);
 
 	/* When the !RCS engines idle waiting upon a semaphore, they lose their
 	 * pagetables and we must reload them before executing the batch.
 	 * We do this on the i915_switch_context() following the wait and
 	 * before the dispatch.
 	 */
-	ppgtt = wait->ctx->ppgtt;
-	if (ppgtt && wait->engine->id != RCS)
-		ppgtt->pd_dirty_rings |= intel_engine_flag(wait->engine);
+	ppgtt = req->ctx->ppgtt;
+	if (ppgtt && req->engine->id != RCS)
+		ppgtt->pd_dirty_rings |= intel_engine_flag(req->engine);
 	return 0;
 }
 
 static int
-gen6_ring_sync(struct drm_i915_gem_request *wait,
-	       struct drm_i915_gem_request *signal)
+gen6_ring_sync_to(struct drm_i915_gem_request *req,
+		  struct drm_i915_gem_request *signal)
 {
-	struct intel_ring *waiter = wait->ring;
+	struct intel_ring *ring = req->ring;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
-	u32 wait_mbox = signal->engine->semaphore.mbox.wait[wait->engine->id];
+	u32 wait_mbox = signal->engine->semaphore.mbox.wait[req->engine->id];
 	int ret;
 
 	WARN_ON(wait_mbox == MI_SEMAPHORE_SYNC_INVALID);
 
-	ret = intel_ring_begin(wait, 4);
+	ret = intel_ring_begin(req, 4);
 	if (ret)
 		return ret;
 
-	intel_ring_emit(waiter, dw1 | wait_mbox);
+	intel_ring_emit(ring, dw1 | wait_mbox);
 	/* Throughout all of the GEM code, seqno passed implies our current
 	 * seqno is >= the last seqno executed. However for hardware the
 	 * comparison is strictly greater than.
 	 */
-	intel_ring_emit(waiter, signal->fence.seqno - 1);
-	intel_ring_emit(waiter, 0);
-	intel_ring_emit(waiter, MI_NOOP);
-	intel_ring_advance(waiter);
+	intel_ring_emit(ring, signal->fence.seqno - 1);
+	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
 	return 0;
 }
@@ -2684,7 +2682,7 @@ static void intel_ring_init_semaphores(struct drm_i915_private *dev_priv,
 	if (INTEL_GEN(dev_priv) >= 8) {
 		u64 offset = i915_gem_obj_ggtt_offset(dev_priv->semaphore_obj);
 
-		engine->semaphore.sync_to = gen8_ring_sync;
+		engine->semaphore.sync_to = gen8_ring_sync_to;
 		engine->semaphore.signal = gen8_xcs_signal;
 
 		for (i = 0; i < I915_NUM_ENGINES; i++) {
@@ -2698,7 +2696,7 @@ static void intel_ring_init_semaphores(struct drm_i915_private *dev_priv,
 			engine->semaphore.signal_ggtt[i] = ring_offset;
 		}
 	} else if (INTEL_GEN(dev_priv) >= 6) {
-		engine->semaphore.sync_to = gen6_ring_sync;
+		engine->semaphore.sync_to = gen6_ring_sync_to;
 		engine->semaphore.signal = gen6_signal;
 
 		/*
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 7a4cf4a14c1a..e59162d89026 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -275,9 +275,9 @@ struct intel_engine_cs {
 		};
 
 		/* AKA wait() */
-		int	(*sync_to)(struct drm_i915_gem_request *to,
-				   struct drm_i915_gem_request *from);
-		int	(*signal)(struct drm_i915_gem_request *signaller_req);
+		int	(*sync_to)(struct drm_i915_gem_request *req,
+				   struct drm_i915_gem_request *signal);
+		int	(*signal)(struct drm_i915_gem_request *req);
 	} semaphore;
 
 	/* Execlists */
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 02/31] drm/i915: Prefer list_first_entry_or_null

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> list_first_entry_or_null() can generate better code than using
> if (!list_empty()) {ptr = list_first_entry()) ..., so put it to use.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

It looks cleaner, not so sure of the code impact.

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/i915_gem_gtt.c      | 12 +++++-------
>  drivers/gpu/drm/i915/i915_gem_request.c  |  8 +++-----
>  drivers/gpu/drm/i915/i915_gem_shrinker.c |  9 +++++----
>  3 files changed, 13 insertions(+), 16 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
> index 30da543e1bdf..38e7d992a20d 100644
> --- a/drivers/gpu/drm/i915/i915_gem_gtt.c
> +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
> @@ -2736,13 +2736,11 @@ static void i915_gtt_color_adjust(struct drm_mm_node *node,
>  	if (node->color != color)
>  		*start += 4096;
>  
> -	if (!list_empty(&node->node_list)) {
> -		node = list_entry(node->node_list.next,
> -				  struct drm_mm_node,
> -				  node_list);
> -		if (node->allocated && node->color != color)
> -			*end -= 4096;
> -	}
> +	node = list_first_entry_or_null(&node->node_list,
> +					struct drm_mm_node,
> +					node_list);
> +	if (node && node->allocated && node->color != color)
> +		*end -= 4096;
>  }
>  
>  static int i915_gem_setup_global_gtt(struct drm_device *dev,
> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 60a3a343b3a8..49396b895a36 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -317,12 +317,10 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>  		return ret;
>  
>  	/* Move the oldest request to the slab-cache (if not in use!) */
> -	if (!list_empty(&engine->request_list)) {
> -		req = list_first_entry(&engine->request_list,
> +	req = list_first_entry_or_null(&engine->request_list,
>  				       typeof(*req), list);
> -		if (i915_gem_request_completed(req))
> -			i915_gem_request_retire(req);
> -	}
> +	if (req && i915_gem_request_completed(req))
> +		i915_gem_request_retire(req);
>  
>  	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
>  	if (!req)
> diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
> index afaa2597e35e..5d4772c146b1 100644
> --- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
> +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
> @@ -163,17 +163,18 @@ i915_gem_shrink(struct drm_i915_private *dev_priv,
>  	 */
>  	for (phase = phases; phase->list; phase++) {
>  		struct list_head still_in_list;
> +		struct drm_i915_gem_object *obj;
>  
>  		if ((flags & phase->bit) == 0)
>  			continue;
>  
>  		INIT_LIST_HEAD(&still_in_list);
> -		while (count < target && !list_empty(phase->list)) {
> -			struct drm_i915_gem_object *obj;
> +		while (count < target &&
> +		       (obj = list_first_entry_or_null(phase->list,
> +						       typeof(*obj),
> +						       global_list))) {
>  			struct i915_vma *vma, *v;
>  
> -			obj = list_first_entry(phase->list,
> -					       typeof(*obj), global_list);
>  			list_move_tail(&obj->global_list, &still_in_list);
>  
>  			if (flags & I915_SHRINK_PURGEABLE &&
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 02/31] drm/i915: Prefer list_first_entry_or_null

[Chris Wilson]

On Mon, Jul 25, 2016 at 10:55:24AM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > list_first_entry_or_null() can generate better code than using
> > if (!list_empty()) {ptr = list_first_entry()) ..., so put it to use.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> 
> It looks cleaner, not so sure of the code impact.

After the patch I sent to linux-kernel@ to improve code gneration for
list_first_entry_or_null() is applied we save a few bytes on each
invocation.

There are a few other places where I could have used
list_first_entry_or_null() but I have already sent patches transforming
the code so I didn't bother with the intermediate minor improvement.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 03/31] drm/i915: Only clear the client pointer when tearing down the file

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Upon release of the file (i.e. the user calls close(fd)), we decouple
> all objects from the client list so that we don't chase the dangling
> file_priv. As we always inspect file_priv first, we only need to nullify
> that pointer and can safely ignore the list_head.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/i915_gem.c | 10 ++--------
>  1 file changed, 2 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index c8436639b3ed..3730aecc1eae 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -4745,21 +4745,15 @@ int i915_gem_freeze_late(struct drm_i915_private *dev_priv)
>  void i915_gem_release(struct drm_device *dev, struct drm_file *file)
>  {
>  	struct drm_i915_file_private *file_priv = file->driver_priv;
> +	struct drm_i915_gem_request *request;
>  
>  	/* Clean up our request list when the client is going away, so that
>  	 * later retire_requests won't dereference our soon-to-be-gone
>  	 * file_priv.
>  	 */
>  	spin_lock(&file_priv->mm.lock);
> -	while (!list_empty(&file_priv->mm.request_list)) {
> -		struct drm_i915_gem_request *request;
> -
> -		request = list_first_entry(&file_priv->mm.request_list,
> -					   struct drm_i915_gem_request,
> -					   client_list);
> -		list_del(&request->client_list);
> +	list_for_each_entry(request, &file_priv->mm.request_list, client_list)
>  		request->file_priv = NULL;
> -	}
>  	spin_unlock(&file_priv->mm.lock);
>  
>  	if (!list_empty(&file_priv->rps.link)) {
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Ro.CI.BAT: success for series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()

[Patchwork]

== Series Details ==

Series: series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()
URL   : https://patchwork.freedesktop.org/series/10230/
State : success

== Summary ==

Series 10230v1 Series without cover letter
http://patchwork.freedesktop.org/api/1.0/series/10230/revisions/1/mbox


ro-bdw-i5-5250u  total:244  pass:219  dwarn:4   dfail:0   fail:8   skip:13 
ro-bdw-i7-5557U  total:244  pass:220  dwarn:3   dfail:0   fail:8   skip:13 
ro-bdw-i7-5600u  total:244  pass:204  dwarn:0   dfail:0   fail:8   skip:32 
ro-bsw-n3050     total:218  pass:173  dwarn:0   dfail:0   fail:2   skip:42 
ro-byt-n2820     total:244  pass:197  dwarn:0   dfail:0   fail:9   skip:38 
ro-hsw-i3-4010u  total:244  pass:212  dwarn:0   dfail:0   fail:8   skip:24 
ro-hsw-i7-4770r  total:244  pass:212  dwarn:0   dfail:0   fail:8   skip:24 
ro-ilk-i7-620lm  total:244  pass:172  dwarn:0   dfail:0   fail:9   skip:63 
ro-ilk1-i5-650   total:239  pass:172  dwarn:0   dfail:0   fail:9   skip:58 
ro-ivb-i7-3770   total:244  pass:203  dwarn:0   dfail:0   fail:8   skip:33 
ro-skl3-i5-6260u total:244  pass:224  dwarn:0   dfail:0   fail:8   skip:12 
ro-snb-i7-2620M  total:244  pass:193  dwarn:0   dfail:0   fail:9   skip:42 

Results at /archive/results/CI_IGT_test/RO_Patchwork_1595/

5c9e3d9 drm-intel-nightly: 2016y-07m-25d-06h-32m-37s UTC integration manifest
9a8596c drm/i915: Rename engine->semaphore.sync_to, engine->sempahore.signal locals
982d63f drm/i915: Simplify calling engine->sync_to
156ff98 drm/i915: Unify legacy/execlists submit_execbuf callbacks
abba4d6 drm/i915: Refactor golden render state emission to unconfuse gcc
2beb0bd drm/i915: Remove duplicate golden render state init from execlists
d646136 drm/i915/ringbuffer: Specialise SNB+ request emission for semaphores
a2e9dc8 drm/i915: Reuse legacy breadcrumbs + tail emission
7a112f0 drm/i915: Stop passing caller's num_dwords to engine->semaphore.signal()
a111a6f drm/i915/lrc: Update function names to match request flow
b82813f drm/i915: Unify request submission
51261a7 drm/i915: Convert engine->write_tail to operate on a request
897e209 drm/i915: Remove intel_ring_get_tail()
e244937 drm/i915: Unify legacy/execlists emission of MI_BATCHBUFFER_START
3e62680 drm/i915: Simplify request_alloc by returning the allocated request
05b985d drm/i915: Remove obsolete engine->gpu_caches_dirty
6dd5c52 drm/i915: Rename intel_pin_and_map_ring()
1a7d6b0 drm/i915: Rename residual ringbuf parameters
d4f682c drm/i915: Rename struct intel_ringbuffer to struct intel_ring
0b6c47a drm/i915: Rename intel_context[engine].ringbuf
7fa3725 drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs
0729469 drm/i915: Rename request->ringbuf to request->ring
30e1fb8 drm/i915: Unify intel_logical_ring_emit and intel_ring_emit
f206a14 drm/i915: Update a couple of hangcheck comments to talk about engines
f961e19 drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c
89ce89b drm/i915: Avoid using intel_engine_cs *ring for GPU error capture
0ccd3f9 drm/i915: Use engine to refer to the user's BSD intel_engine_cs
92e3bae drm/i915/cmdparser: Remove stray intel_engine_cs *ring
c3c9e9b drm/i915: Only drop the batch-pool's object reference
496a446 drm/i915: Only clear the client pointer when tearing down the file
e3eefd2 drm/i915: Prefer list_first_entry_or_null
aab8747 drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 04/31] drm/i915: Only drop the batch-pool's object reference

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> The obj->batch_pool_link is only inspected when traversing the batch
> pool list and when on the batch pool list the object is referenced. Thus
> when freeing the batch pool list, we only need to unreference the object
> and do not have to worry about the obj->batch_pool_link.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_gem_batch_pool.c | 13 ++++++-------
>  1 file changed, 6 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> index 3507b2753fd3..825981b5aa40 100644
> --- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> +++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> @@ -68,15 +68,14 @@ void i915_gem_batch_pool_fini(struct i915_gem_batch_pool *pool)
>  	WARN_ON(!mutex_is_locked(&pool->dev->struct_mutex));
>  
>  	for (n = 0; n < ARRAY_SIZE(pool->cache_list); n++) {
> -		while (!list_empty(&pool->cache_list[n])) {
> -			struct drm_i915_gem_object *obj =
> -				list_first_entry(&pool->cache_list[n],
> -						 struct drm_i915_gem_object,
> -						 batch_pool_link);
> +		struct drm_i915_gem_object *obj, *next;
>  
> -			list_del(&obj->batch_pool_link);
> +		list_for_each_entry_safe(obj, next,

Why _safe? i915_gem_free_object does not touch the batch_pool_link.

Regards, Joonas

> +					 &pool->cache_list[n],
> +					 batch_pool_link)
>  			i915_gem_object_put(obj);
> -		}
> +
> +		INIT_LIST_HEAD(&pool->cache_list[n]);
>  	}
>  }
>  
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 05/31] drm/i915/cmdparser: Remove stray intel_engine_cs *ring

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> When we refer to intel_engine_cs, we want to use engine so as not to
> confuse ourselves about ringbuffers.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_cmd_parser.c  |  4 ++--
>  drivers/gpu/drm/i915/i915_drv.h         |  5 +++--
>  drivers/gpu/drm/i915/intel_ringbuffer.h | 10 +++++-----
>  3 files changed, 10 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
> index b0fd6a7b0603..b5b520176c59 100644
> --- a/drivers/gpu/drm/i915/i915_cmd_parser.c
> +++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
> @@ -636,7 +636,7 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
>  	return ret;
>  }
>  
> -static bool check_sorted(int ring_id,
> +static bool check_sorted(int engine_id,
>  			 const struct drm_i915_reg_descriptor *reg_table,
>  			 int reg_count)
>  {
> @@ -649,7 +649,7 @@ static bool check_sorted(int ring_id,
>  
>  		if (curr < previous) {
>  			DRM_ERROR("CMD: table not sorted ring=%d entry=%d reg=0x%08X prev=0x%08X\n",

How about the message?

> -				  ring_id, i, curr, previous);
> +				  engine_id, i, curr, previous);
>  			ret = false;
>  		}
>  
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index 9f655e2add66..d4c5c2edf252 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -2500,8 +2500,9 @@ struct drm_i915_cmd_descriptor {
>  /*
>   * A table of commands requiring special handling by the command parser.
>   *
> - * Each ring has an array of tables. Each table consists of an array of command
> - * descriptors, which must be sorted with command opcodes in ascending order.
> + * Each engine has an array of tables. Each table consists of an array of
> + * command descriptors, which must be sorted with command opcodes in
> + * ascending order.
>   */
>  struct drm_i915_cmd_table {
>  	const struct drm_i915_cmd_descriptor *table;
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 0f8019488d33..0a830c116886 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -340,7 +340,7 @@ struct intel_engine_cs {
>  
>  	/*
>  	 * Table of commands the command parser needs to know about
> -	 * for this ring.
> +	 * for this engine.
>  	 */
>  	DECLARE_HASHTABLE(cmd_hash, I915_CMD_HASH_ORDER);
>  
> @@ -354,11 +354,11 @@ struct intel_engine_cs {
>  	 * Returns the bitmask for the length field of the specified command.
>  	 * Return 0 for an unrecognized/invalid command.
>  	 *
> -	 * If the command parser finds an entry for a command in the ring's
> +	 * If the command parser finds an entry for a command in the engines's

s/engines's/engine's/

>  	 * cmd_tables, it gets the command's length based on the table entry.
> -	 * If not, it calls this function to determine the per-ring length field
> -	 * encoding for the command (i.e. certain opcode ranges use certain bits
> -	 * to encode the command length in the header).
> +	 * If not, it calls this function to determine the per-engine length
> +	 * field encoding for the command (i.e. different opcode ranges use
> +	 * certain bits to encode the command length in the header).
>  	 */
>  	u32 (*get_cmd_length_mask)(u32 cmd_header);
>  };
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 06/31] drm/i915: Use engine to refer to the user's BSD intel_engine_cs

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> This patch transitions the execbuf engine selection away from using the
> ring nomenclature - though we still refer to the user's incoming
> selector as their user_ring_id.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/i915_drv.h            |  4 ++--
>  drivers/gpu/drm/i915/i915_gem.c            |  2 +-
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c | 15 ++++++++-------
>  3 files changed, 11 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index d4c5c2edf252..b686b741ea87 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -402,7 +402,7 @@ struct drm_i915_file_private {
>  		unsigned boosts;
>  	} rps;
>  
> -	unsigned int bsd_ring;
> +	unsigned int bsd_engine;
>  };
>  
>  /* Used by dp and fdi links */
> @@ -1331,7 +1331,7 @@ struct i915_gem_mm {
>  	bool interruptible;
>  
>  	/* the indicator for dispatch video commands on two BSD rings */
> -	unsigned int bsd_ring_dispatch_index;
> +	unsigned int bsd_engine_dispatch_index;
>  
>  	/** Bit 6 swizzling required for X tiling */
>  	uint32_t bit_6_swizzle_x;
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 3730aecc1eae..e155e8dd28ed 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -4782,7 +4782,7 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file)
>  	spin_lock_init(&file_priv->mm.lock);
>  	INIT_LIST_HEAD(&file_priv->mm.request_list);
>  
> -	file_priv->bsd_ring = -1;
> +	file_priv->bsd_engine = -1;
>  
>  	ret = i915_gem_context_open(dev, file);
>  	if (ret)
> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> index f8d8ae39fc2f..9ef68e86b483 100644
> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> @@ -1351,23 +1351,24 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
>  
>  /**
>   * Find one BSD ring to dispatch the corresponding BSD command.
> - * The ring index is returned.
> + * The engine index is returned.
>   */
>  static unsigned int
> -gen8_dispatch_bsd_ring(struct drm_i915_private *dev_priv, struct drm_file *file)
> +gen8_dispatch_bsd_engine(struct drm_i915_private *dev_priv,
> +			 struct drm_file *file)
>  {
>  	struct drm_i915_file_private *file_priv = file->driver_priv;
>  
>  	/* Check whether the file_priv has already selected one ring. */
> -	if ((int)file_priv->bsd_ring < 0) {
> +	if ((int)file_priv->bsd_engine < 0) {
>  		/* If not, use the ping-pong mechanism to select one. */
>  		mutex_lock(&dev_priv->drm.struct_mutex);
> -		file_priv->bsd_ring = dev_priv->mm.bsd_ring_dispatch_index;
> -		dev_priv->mm.bsd_ring_dispatch_index ^= 1;
> +		file_priv->bsd_engine = dev_priv->mm.bsd_engine_dispatch_index;
> +		dev_priv->mm.bsd_engine_dispatch_index ^= 1;
>  		mutex_unlock(&dev_priv->drm.struct_mutex);
>  	}
>  
> -	return file_priv->bsd_ring;
> +	return file_priv->bsd_engine;
>  }
>  
>  #define I915_USER_RINGS (4)
> @@ -1404,7 +1405,7 @@ eb_select_engine(struct drm_i915_private *dev_priv,
>  		unsigned int bsd_idx = args->flags & I915_EXEC_BSD_MASK;
>  
>  		if (bsd_idx == I915_EXEC_BSD_DEFAULT) {
> -			bsd_idx = gen8_dispatch_bsd_ring(dev_priv, file);
> +			bsd_idx = gen8_dispatch_bsd_engine(dev_priv, file);
>  		} else if (bsd_idx >= I915_EXEC_BSD_RING1 &&
>  			   bsd_idx <= I915_EXEC_BSD_RING2) {
>  			bsd_idx >>= I915_EXEC_BSD_SHIFT;
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 04/31] drm/i915: Only drop the batch-pool's object reference

[Chris Wilson]

On Mon, Jul 25, 2016 at 11:38:07AM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > The obj->batch_pool_link is only inspected when traversing the batch
> > pool list and when on the batch pool list the object is referenced. Thus
> > when freeing the batch pool list, we only need to unreference the object
> > and do not have to worry about the obj->batch_pool_link.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > ---
> >  drivers/gpu/drm/i915/i915_gem_batch_pool.c | 13 ++++++-------
> >  1 file changed, 6 insertions(+), 7 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > index 3507b2753fd3..825981b5aa40 100644
> > --- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > +++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > @@ -68,15 +68,14 @@ void i915_gem_batch_pool_fini(struct i915_gem_batch_pool *pool)
> >  	WARN_ON(!mutex_is_locked(&pool->dev->struct_mutex));
> >  
> >  	for (n = 0; n < ARRAY_SIZE(pool->cache_list); n++) {
> > -		while (!list_empty(&pool->cache_list[n])) {
> > -			struct drm_i915_gem_object *obj =
> > -				list_first_entry(&pool->cache_list[n],
> > -						 struct drm_i915_gem_object,
> > -						 batch_pool_link);
> > +		struct drm_i915_gem_object *obj, *next;
> >  
> > -			list_del(&obj->batch_pool_link);
> > +		list_for_each_entry_safe(obj, next,
> 
> Why _safe? i915_gem_free_object does not touch the batch_pool_link.

Not directly, but it does free the memory including the batch_pool_link. :)
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 08/31] drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> A few places we use ring when referring to the struct intel_engine_cs. An
> anachronism we are pruning out.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_gem.c | 24 ++++++++++++------------
>  1 file changed, 12 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index e155e8dd28ed..7bfce1d5c61b 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -46,7 +46,7 @@ static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *o
>  static void
>  i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
>  static void
> -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring);
> +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int engine);
>  
>  static bool cpu_cache_is_coherent(struct drm_device *dev,
>  				  enum i915_cache_level level)
> @@ -1385,10 +1385,10 @@ static void
>  i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
>  			       struct drm_i915_gem_request *req)
>  {
> -	int ring = req->engine->id;
> +	int idx = req->engine->id;

See below.

>  
> -	if (obj->last_read_req[ring] == req)
> -		i915_gem_object_retire__read(obj, ring);
> +	if (obj->last_read_req[idx] == req)
> +		i915_gem_object_retire__read(obj, idx);
>  	else if (obj->last_write_req == req)
>  		i915_gem_object_retire__write(obj);
>  
> @@ -2381,20 +2381,20 @@ i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
>  }
>  
>  static void
> -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
> +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)

I do not fancy declaring different variable names than are used. Also,
idx is very cryptic in this function signature (one would think of
object index).

Regards, Joonas

>  {
>  	struct i915_vma *vma;
>  
> -	GEM_BUG_ON(obj->last_read_req[ring] == NULL);
> -	GEM_BUG_ON(!(obj->active & (1 << ring)));
> +	GEM_BUG_ON(obj->last_read_req[idx] == NULL);
> +	GEM_BUG_ON(!(obj->active & (1 << idx)));
>  
> -	list_del_init(&obj->engine_list[ring]);
> -	i915_gem_request_assign(&obj->last_read_req[ring], NULL);
> +	list_del_init(&obj->engine_list[idx]);
> +	i915_gem_request_assign(&obj->last_read_req[idx], NULL);
>  
> -	if (obj->last_write_req && obj->last_write_req->engine->id == ring)
> +	if (obj->last_write_req && obj->last_write_req->engine->id == idx)
>  		i915_gem_object_retire__write(obj);
>  
> -	obj->active &= ~(1 << ring);
> +	obj->active &= ~(1 << idx);
>  	if (obj->active)
>  		return;
>  
> @@ -4599,7 +4599,7 @@ int i915_gem_init(struct drm_device *dev)
>  
>  	ret = i915_gem_init_hw(dev);
>  	if (ret == -EIO) {
> -		/* Allow ring initialisation to fail by marking the GPU as
> +		/* Allow engine initialisation to fail by marking the GPU as
>  		 * wedged. But we only want to do this where the GPU is angry,
>  		 * for all other failure, such as an allocation failure, bail.
>  		 */
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 09/31] drm/i915: Update a couple of hangcheck comments to talk about engines

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> We still have lots of comments that refer to the old ring when we mean
> struct intel_engine_cs and its hardware correspondence. This patch fixes
> an instance inside hangcheck to talk about engines.
> 

These could be squashed down a bit too. 

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_irq.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
> index 7104dc1463eb..f5bf4f913a91 100644
> --- a/drivers/gpu/drm/i915/i915_irq.c
> +++ b/drivers/gpu/drm/i915/i915_irq.c
> @@ -3140,13 +3140,13 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
>  				}
>  			} else {
>  				/* We always increment the hangcheck score
> -				 * if the ring is busy and still processing
> +				 * if the engine is busy and still processing
>  				 * the same request, so that no single request
>  				 * can run indefinitely (such as a chain of
>  				 * batches). The only time we do not increment
>  				 * the hangcheck score on this ring, if this
> -				 * ring is in a legitimate wait for another
> -				 * ring. In that case the waiting ring is a
> +				 * engine is in a legitimate wait for another
> +				 * engine. In that case the waiting engine is a
>  				 * victim and we want to be sure we catch the
>  				 * right culprit. Then every time we do kick
>  				 * the ring, add a small increment to the
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 08/31] drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c

[Chris Wilson]

On Mon, Jul 25, 2016 at 11:45:42AM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > A few places we use ring when referring to the struct intel_engine_cs. An
> > anachronism we are pruning out.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > ---
> >  drivers/gpu/drm/i915/i915_gem.c | 24 ++++++++++++------------
> >  1 file changed, 12 insertions(+), 12 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> > index e155e8dd28ed..7bfce1d5c61b 100644
> > --- a/drivers/gpu/drm/i915/i915_gem.c
> > +++ b/drivers/gpu/drm/i915/i915_gem.c
> > @@ -46,7 +46,7 @@ static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *o
> >  static void
> >  i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
> >  static void
> > -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring);
> > +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int engine);
> >  
> >  static bool cpu_cache_is_coherent(struct drm_device *dev,
> >  				  enum i915_cache_level level)
> > @@ -1385,10 +1385,10 @@ static void
> >  i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
> >  			       struct drm_i915_gem_request *req)
> >  {
> > -	int ring = req->engine->id;
> > +	int idx = req->engine->id;
> 
> See below.
> 
> >  
> > -	if (obj->last_read_req[ring] == req)
> > -		i915_gem_object_retire__read(obj, ring);
> > +	if (obj->last_read_req[idx] == req)
> > +		i915_gem_object_retire__read(obj, idx);
> >  	else if (obj->last_write_req == req)
> >  		i915_gem_object_retire__write(obj);
> >  
> > @@ -2381,20 +2381,20 @@ i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
> >  }
> >  
> >  static void
> > -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
> > +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)
> 
> I do not fancy declaring different variable names than are used. Also,
> idx is very cryptic in this function signature (one would think of
> object index).

If you look at the later patches (posted earlier ;) using index is the
less cryptic option as it really does refer to the index of the tracker.

There was a desired to avoid using ring and here I was trying to avoid
confusion with activity tracking.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 12/31] drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Having ringbuf->ring point to an engine is confusing, so rename it once
> again to ring->engine.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

I still do not see a connection between commit message and content, so
you could mentione renaming the function (which is all this patch
does).

For code,
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 12 ++++++------
>  1 file changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 7ba31724feaf..c4582b9a6862 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -2186,7 +2186,7 @@ static void intel_ring_context_unpin(struct i915_gem_context *ctx,
>  	i915_gem_context_put(ctx);
>  }
>  
> -static int intel_init_ring_buffer(struct intel_engine_cs *engine)
> +static int intel_init_engine(struct intel_engine_cs *engine)
>  {
>  	struct drm_i915_private *dev_priv = engine->i915;
>  	struct intel_ringbuffer *ringbuf;
> @@ -2883,7 +2883,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
>  	engine->init_hw = init_render_ring;
>  	engine->cleanup = render_ring_cleanup;
>  
> -	ret = intel_init_ring_buffer(engine);
> +	ret = intel_init_engine(engine);
>  	if (ret)
>  		return ret;
>  
> @@ -2922,7 +2922,7 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
>  			engine->irq_enable_mask = I915_BSD_USER_INTERRUPT;
>  	}
>  
> -	return intel_init_ring_buffer(engine);
> +	return intel_init_engine(engine);
>  }
>  
>  /**
> @@ -2936,7 +2936,7 @@ int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine)
>  
>  	engine->flush = gen6_bsd_ring_flush;
>  
> -	return intel_init_ring_buffer(engine);
> +	return intel_init_engine(engine);
>  }
>  
>  int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
> @@ -2949,7 +2949,7 @@ int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
>  	if (INTEL_GEN(dev_priv) < 8)
>  		engine->irq_enable_mask = GT_BLT_USER_INTERRUPT;
>  
> -	return intel_init_ring_buffer(engine);
> +	return intel_init_engine(engine);
>  }
>  
>  int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
> @@ -2966,7 +2966,7 @@ int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine)
>  		engine->irq_disable = hsw_vebox_irq_disable;
>  	}
>  
> -	return intel_init_ring_buffer(engine);
> +	return intel_init_engine(engine);
>  }
>  
>  int
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 15/31] drm/i915: Rename residual ringbuf parameters

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Now that we have a clear ring/engine split and a struct intel_ring, we
> no longer need the stopgap ringbuf names.

+1 vote for squashing related renames, should be make picking them
later easier, unless there is huge pieces of code in between.

> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 66 ++++++++++++++++-----------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h |  6 +--
>  2 files changed, 36 insertions(+), 36 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 90572b862eec..5ff44eb46f55 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -47,15 +47,15 @@ int __intel_ring_space(int head, int tail, int size)
>  	return space - I915_RING_FREE_SPACE;
>  }
>  
> -void intel_ring_update_space(struct intel_ring *ringbuf)
> +void intel_ring_update_space(struct intel_ring *ring)
>  {
> -	if (ringbuf->last_retired_head != -1) {
> -		ringbuf->head = ringbuf->last_retired_head;
> -		ringbuf->last_retired_head = -1;
> +	if (ring->last_retired_head != -1) {
> +		ring->head = ring->last_retired_head;
> +		ring->last_retired_head = -1;
>  	}
>  
> -	ringbuf->space = __intel_ring_space(ringbuf->head & HEAD_ADDR,
> -					    ringbuf->tail, ringbuf->size);
> +	ring->space = __intel_ring_space(ring->head & HEAD_ADDR,
> +					 ring->tail, ring->size);

See, no getter function used here either for ring->tail and how clean
it looks.

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH v2] drm/i915/cmdparser: Remove stray intel_engine_cs *ring

[Chris Wilson]

When we refer to intel_engine_cs, we want to use engine so as not to
confuse ourselves about ringbuffers.

v2: Rename all the functions as well, as well as a few more stray comments.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: http://patchwork.freedesktop.org/patch/msgid/1469432687-22756-6-git-send-email-chris@chris-wilson.co.uk
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_cmd_parser.c     | 72 +++++++++++++++---------------
 drivers/gpu/drm/i915/i915_drv.h            | 23 +++++-----
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 14 +++---
 drivers/gpu/drm/i915/intel_engine_cs.c     |  2 +-
 drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c    |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.h    | 10 ++---
 7 files changed, 64 insertions(+), 61 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index b0fd6a7..8db144b 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -62,23 +62,23 @@
  * The parser always rejects such commands.
  *
  * The majority of the problematic commands fall in the MI_* range, with only a
- * few specific commands on each ring (e.g. PIPE_CONTROL and MI_FLUSH_DW).
+ * few specific commands on each engine (e.g. PIPE_CONTROL and MI_FLUSH_DW).
  *
  * Implementation:
- * Each ring maintains tables of commands and registers which the parser uses in
- * scanning batch buffers submitted to that ring.
+ * Each engine maintains tables of commands and registers which the parser
+ * uses in scanning batch buffers submitted to that engine.
  *
  * Since the set of commands that the parser must check for is significantly
  * smaller than the number of commands supported, the parser tables contain only
  * those commands required by the parser. This generally works because command
  * opcode ranges have standard command length encodings. So for commands that
  * the parser does not need to check, it can easily skip them. This is
- * implemented via a per-ring length decoding vfunc.
+ * implemented via a per-engine length decoding vfunc.
  *
  * Unfortunately, there are a number of commands that do not follow the standard
  * length encoding for their opcode range, primarily amongst the MI_* commands.
  * To handle this, the parser provides a way to define explicit "skip" entries
- * in the per-ring command tables.
+ * in the per-engine command tables.
  *
  * Other command table entries map fairly directly to high level categories
  * mentioned above: rejected, master-only, register whitelist. The parser
@@ -603,7 +603,7 @@ static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
 	return 0;
 }
 
-static bool validate_cmds_sorted(struct intel_engine_cs *engine,
+static bool validate_cmds_sorted(const struct intel_engine_cs *engine,
 				 const struct drm_i915_cmd_table *cmd_tables,
 				 int cmd_table_count)
 {
@@ -624,8 +624,9 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
 			u32 curr = desc->cmd.value & desc->cmd.mask;
 
 			if (curr < previous) {
-				DRM_ERROR("CMD: table not sorted ring=%d table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
-					  engine->id, i, j, curr, previous);
+				DRM_ERROR("CMD: %s [%d] command table not sorted: table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
+					  engine->name, engine->id,
+					  i, j, curr, previous);
 				ret = false;
 			}
 
@@ -636,7 +637,7 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
 	return ret;
 }
 
-static bool check_sorted(int ring_id,
+static bool check_sorted(const struct intel_engine_cs *engine,
 			 const struct drm_i915_reg_descriptor *reg_table,
 			 int reg_count)
 {
@@ -648,8 +649,9 @@ static bool check_sorted(int ring_id,
 		u32 curr = i915_mmio_reg_offset(reg_table[i].addr);
 
 		if (curr < previous) {
-			DRM_ERROR("CMD: table not sorted ring=%d entry=%d reg=0x%08X prev=0x%08X\n",
-				  ring_id, i, curr, previous);
+			DRM_ERROR("CMD: engine %s [%d] register table not sorted: entry=%d reg=0x%08X prev=0x%08X\n",
+				  engine->name, engine->id,
+				  i, curr, previous);
 			ret = false;
 		}
 
@@ -666,7 +668,7 @@ static bool validate_regs_sorted(struct intel_engine_cs *engine)
 
 	for (i = 0; i < engine->reg_table_count; i++) {
 		table = &engine->reg_tables[i];
-		if (!check_sorted(engine->id, table->regs, table->num_regs))
+		if (!check_sorted(engine, table->regs, table->num_regs))
 			return false;
 	}
 
@@ -736,7 +738,7 @@ static void fini_hash_table(struct intel_engine_cs *engine)
 }
 
 /**
- * i915_cmd_parser_init_ring() - set cmd parser related fields for a ringbuffer
+ * intel_engine_init_cmd_parser() - set cmd parser related fields for an engine
  * @engine: the engine to initialize
  *
  * Optionally initializes fields related to batch buffer command parsing in the
@@ -745,7 +747,7 @@ static void fini_hash_table(struct intel_engine_cs *engine)
  *
  * Return: non-zero if initialization fails
  */
-int i915_cmd_parser_init_ring(struct intel_engine_cs *engine)
+int intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
 {
 	const struct drm_i915_cmd_table *cmd_tables;
 	int cmd_table_count;
@@ -806,8 +808,7 @@ int i915_cmd_parser_init_ring(struct intel_engine_cs *engine)
 		engine->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
 		break;
 	default:
-		DRM_ERROR("CMD: cmd_parser_init with unknown ring: %d\n",
-			  engine->id);
+		MISSING_CASE(engine->id);
 		BUG();
 	}
 
@@ -829,13 +830,13 @@ int i915_cmd_parser_init_ring(struct intel_engine_cs *engine)
 }
 
 /**
- * i915_cmd_parser_fini_ring() - clean up cmd parser related fields
+ * intel_engine_cleanup_cmd_parser() - clean up cmd parser related fields
  * @engine: the engine to clean up
  *
  * Releases any resources related to command parsing that may have been
- * initialized for the specified ring.
+ * initialized for the specified engine.
  */
-void i915_cmd_parser_fini_ring(struct intel_engine_cs *engine)
+void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine)
 {
 	if (!engine->needs_cmd_parser)
 		return;
@@ -866,9 +867,9 @@ find_cmd_in_table(struct intel_engine_cs *engine,
  * Returns a pointer to a descriptor for the command specified by cmd_header.
  *
  * The caller must supply space for a default descriptor via the default_desc
- * parameter. If no descriptor for the specified command exists in the ring's
+ * parameter. If no descriptor for the specified command exists in the engine's
  * command parser tables, this function fills in default_desc based on the
- * ring's default length encoding and returns default_desc.
+ * engine's default length encoding and returns default_desc.
  */
 static const struct drm_i915_cmd_descriptor*
 find_cmd(struct intel_engine_cs *engine,
@@ -1023,15 +1024,16 @@ unpin_src:
 }
 
 /**
- * i915_needs_cmd_parser() - should a given ring use software command parsing?
+ * intel_engine_needs_cmd_parser() - should a given engine use software
+ *                                   command parsing?
  * @engine: the engine in question
  *
  * Only certain platforms require software batch buffer command parsing, and
  * only when enabled via module parameter.
  *
- * Return: true if the ring requires software command parsing
+ * Return: true if the engine requires software command parsing
  */
-bool i915_needs_cmd_parser(struct intel_engine_cs *engine)
+bool intel_engine_needs_cmd_parser(struct intel_engine_cs *engine)
 {
 	if (!engine->needs_cmd_parser)
 		return false;
@@ -1078,8 +1080,8 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 						   reg_addr);
 
 			if (!reg) {
-				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (ring=%d)\n",
-						 reg_addr, *cmd, engine->id);
+				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (exec_id=%d)\n",
+						 reg_addr, *cmd, engine->exec_id);
 				return false;
 			}
 
@@ -1159,11 +1161,11 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 				desc->bits[i].mask;
 
 			if (dword != desc->bits[i].expected) {
-				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (ring=%d)\n",
+				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (exec_id=%d)\n",
 						 *cmd,
 						 desc->bits[i].mask,
 						 desc->bits[i].expected,
-						 dword, engine->id);
+						 dword, engine->exec_id);
 				return false;
 			}
 		}
@@ -1189,12 +1191,12 @@ static bool check_cmd(const struct intel_engine_cs *engine,
  * Return: non-zero if the parser finds violations or otherwise fails; -EACCES
  * if the batch appears legal but should use hardware parsing
  */
-int i915_parse_cmds(struct intel_engine_cs *engine,
-		    struct drm_i915_gem_object *batch_obj,
-		    struct drm_i915_gem_object *shadow_batch_obj,
-		    u32 batch_start_offset,
-		    u32 batch_len,
-		    bool is_master)
+int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+			    struct drm_i915_gem_object *batch_obj,
+			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u32 batch_start_offset,
+			    u32 batch_len,
+			    bool is_master)
 {
 	u32 *cmd, *batch_base, *batch_end;
 	struct drm_i915_cmd_descriptor default_desc = { 0 };
@@ -1295,7 +1297,7 @@ int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv)
 
 	/* If the command parser is not enabled, report 0 - unsupported */
 	for_each_engine(engine, dev_priv) {
-		if (i915_needs_cmd_parser(engine)) {
+		if (intel_engine_needs_cmd_parser(engine)) {
 			active = true;
 			break;
 		}
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 9f655e2..ea9b953 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2500,8 +2500,9 @@ struct drm_i915_cmd_descriptor {
 /*
  * A table of commands requiring special handling by the command parser.
  *
- * Each ring has an array of tables. Each table consists of an array of command
- * descriptors, which must be sorted with command opcodes in ascending order.
+ * Each engine has an array of tables. Each table consists of an array of
+ * command descriptors, which must be sorted with command opcodes in
+ * ascending order.
  */
 struct drm_i915_cmd_table {
 	const struct drm_i915_cmd_descriptor *table;
@@ -3529,15 +3530,15 @@ const char *i915_cache_level_str(struct drm_i915_private *i915, int type);
 
 /* i915_cmd_parser.c */
 int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv);
-int i915_cmd_parser_init_ring(struct intel_engine_cs *engine);
-void i915_cmd_parser_fini_ring(struct intel_engine_cs *engine);
-bool i915_needs_cmd_parser(struct intel_engine_cs *engine);
-int i915_parse_cmds(struct intel_engine_cs *engine,
-		    struct drm_i915_gem_object *batch_obj,
-		    struct drm_i915_gem_object *shadow_batch_obj,
-		    u32 batch_start_offset,
-		    u32 batch_len,
-		    bool is_master);
+int intel_engine_init_cmd_parser(struct intel_engine_cs *engine);
+void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine);
+bool intel_engine_needs_cmd_parser(struct intel_engine_cs *engine);
+int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+			    struct drm_i915_gem_object *batch_obj,
+			    struct drm_i915_gem_object *shadow_batch_obj,
+			    u32 batch_start_offset,
+			    u32 batch_len,
+			    bool is_master);
 
 /* i915_suspend.c */
 extern int i915_save_state(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index f8d8ae3..cd3f873 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1216,12 +1216,12 @@ i915_gem_execbuffer_parse(struct intel_engine_cs *engine,
 	if (IS_ERR(shadow_batch_obj))
 		return shadow_batch_obj;
 
-	ret = i915_parse_cmds(engine,
-			      batch_obj,
-			      shadow_batch_obj,
-			      batch_start_offset,
-			      batch_len,
-			      is_master);
+	ret = intel_engine_cmd_parser(engine,
+				      batch_obj,
+				      shadow_batch_obj,
+				      batch_start_offset,
+				      batch_len,
+				      is_master);
 	if (ret)
 		goto err;
 
@@ -1563,7 +1563,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	}
 
 	params->args_batch_start_offset = args->batch_start_offset;
-	if (i915_needs_cmd_parser(engine) && args->batch_len) {
+	if (intel_engine_needs_cmd_parser(engine) && args->batch_len) {
 		struct drm_i915_gem_object *parsed_batch_obj;
 
 		parsed_batch_obj = i915_gem_execbuffer_parse(engine,
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index f4a35ec..e28873c 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -207,5 +207,5 @@ int intel_engine_init_common(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	return i915_cmd_parser_init_ring(engine);
+	return intel_engine_init_cmd_parser(engine);
 }
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index daf1279..dd3f490 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1925,7 +1925,7 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *engine)
 	if (engine->cleanup)
 		engine->cleanup(engine);
 
-	i915_cmd_parser_fini_ring(engine);
+	intel_engine_cleanup_cmd_parser(engine);
 	i915_gem_batch_pool_fini(&engine->batch_pool);
 
 	intel_engine_fini_breadcrumbs(engine);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 0b5d1de..15acaf6 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -2267,7 +2267,7 @@ void intel_cleanup_engine(struct intel_engine_cs *engine)
 		cleanup_phys_status_page(engine);
 	}
 
-	i915_cmd_parser_fini_ring(engine);
+	intel_engine_cleanup_cmd_parser(engine);
 	i915_gem_batch_pool_fini(&engine->batch_pool);
 	intel_engine_fini_breadcrumbs(engine);
 
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 0f80194..9a0a026 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -340,7 +340,7 @@ struct intel_engine_cs {
 
 	/*
 	 * Table of commands the command parser needs to know about
-	 * for this ring.
+	 * for this engine.
 	 */
 	DECLARE_HASHTABLE(cmd_hash, I915_CMD_HASH_ORDER);
 
@@ -354,11 +354,11 @@ struct intel_engine_cs {
 	 * Returns the bitmask for the length field of the specified command.
 	 * Return 0 for an unrecognized/invalid command.
 	 *
-	 * If the command parser finds an entry for a command in the ring's
+	 * If the command parser finds an entry for a command in the engine's
 	 * cmd_tables, it gets the command's length based on the table entry.
-	 * If not, it calls this function to determine the per-ring length field
-	 * encoding for the command (i.e. certain opcode ranges use certain bits
-	 * to encode the command length in the header).
+	 * If not, it calls this function to determine the per-engine length
+	 * field encoding for the command (i.e. different opcode ranges use
+	 * certain bits to encode the command length in the header).
 	 */
 	u32 (*get_cmd_length_mask)(u32 cmd_header);
 };
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 12/31] drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs

[Chris Wilson]

On Mon, Jul 25, 2016 at 11:49:45AM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > Having ringbuf->ring point to an engine is confusing, so rename it once
> > again to ring->engine.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> 
> I still do not see a connection between commit message and content, so
> you could mentione renaming the function (which is all this patch
> does).

That's weird, I thought I had dropped this patch - it wasn't intended to
be sent again as thinking about how this would look in an
intel_legacy_submission.c says this is pointless churn.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Space for flushing the GPU cache prior to completing the request is
> preallocated and so cannot fail.

Patch title and commit message have some disconnect. Could you explain
in a bit more detail what made gpu_caches_dirty obsolete?

Also, worth mentioning that after this change legacy/execlist flushing
code is unified (could be split patch too? With your "Now that ..."
reasoning).

Somebody not reviewing the series linearly might feel lost.

> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> index ca1d4f573832..048050176ff9 100644
> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> @@ -998,10 +998,8 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
>  	if (flush_domains & I915_GEM_DOMAIN_GTT)
>  		wmb();
>  
> -	/* Unconditionally invalidate gpu caches and ensure that we do flush
> -	 * any residual writes from the previous batch.
> -	 */
> -	return intel_engine_invalidate_all_caches(req);
> +	/* Unconditionally invalidate gpu caches and TLBs. */

A nitpick, but maybe s/gpu/GPU/


> +	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
>  }
>  
>  static bool

<SNIP>

> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 942b5b1f1602..7b772d914e23 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -451,10 +451,9 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  	 * what.
>  	 */
>  	if (flush_caches) {
> -		if (i915.enable_execlists)
> -			ret = logical_ring_flush_all_caches(request);
> -		else
> -			ret = intel_engine_flush_all_caches(request);
> +		ret = request->engine->emit_flush(request,
> +						  0, I915_GEM_GPU_DOMAINS);
> +
>  		/* Not allowed to fail! */
>  		WARN(ret, "*_ring_flush_all_caches failed: %d!\n", ret);

Fix this message too.

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> If is simpler and leads to more readable code through the callstack if
> the allocation returns the allocated struct through the return value.
> 
> The importance of this is that it no longer looks like we accidentally
> allocate requests as side-effect of calling certain functions.
> 

Dave seems to have expressed to wish to review this around January, so
CC'ing him here.

From me,

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_drv.h            |  3 +-
>  drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
>  drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
>  drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
>  drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
>  drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
>  drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
>  8 files changed, 79 insertions(+), 140 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index 09e72b0510ab..1031f5a7126a 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -3171,8 +3171,7 @@ static inline void i915_gem_object_unpin_map(struct drm_i915_gem_object *obj)
>  
>  int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
>  int i915_gem_object_sync(struct drm_i915_gem_object *obj,
> -			 struct intel_engine_cs *to,
> -			 struct drm_i915_gem_request **to_req);
> +			 struct drm_i915_gem_request *to);
>  void i915_vma_move_to_active(struct i915_vma *vma,
>  			     struct drm_i915_gem_request *req);
>  int i915_gem_dumb_create(struct drm_file *file_priv,
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 59890f523c5f..b6c4ff63725f 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -2845,51 +2845,35 @@ out:
>  
>  static int
>  __i915_gem_object_sync(struct drm_i915_gem_object *obj,
> -		       struct intel_engine_cs *to,
> -		       struct drm_i915_gem_request *from_req,
> -		       struct drm_i915_gem_request **to_req)
> +		       struct drm_i915_gem_request *to,
> +		       struct drm_i915_gem_request *from)
>  {
> -	struct intel_engine_cs *from;
>  	int ret;
>  
> -	from = i915_gem_request_get_engine(from_req);
> -	if (to == from)
> +	if (to->engine == from->engine)
>  		return 0;
>  
> -	if (i915_gem_request_completed(from_req))
> +	if (i915_gem_request_completed(from))
>  		return 0;
>  
>  	if (!i915.semaphores) {
> -		struct drm_i915_private *i915 = to_i915(obj->base.dev);
> -		ret = __i915_wait_request(from_req,
> -					  i915->mm.interruptible,
> +		ret = __i915_wait_request(from,
> +					  from->i915->mm.interruptible,
>  					  NULL,
>  					  NO_WAITBOOST);
>  		if (ret)
>  			return ret;
>  
> -		i915_gem_object_retire_request(obj, from_req);
> +		i915_gem_object_retire_request(obj, from);
>  	} else {
> -		int idx = intel_engine_sync_index(from, to);
> -		u32 seqno = i915_gem_request_get_seqno(from_req);
> +		int idx = intel_engine_sync_index(from->engine, to->engine);
> +		u32 seqno = i915_gem_request_get_seqno(from);
>  
> -		WARN_ON(!to_req);
> -
> -		if (seqno <= from->semaphore.sync_seqno[idx])
> +		if (seqno <= from->engine->semaphore.sync_seqno[idx])
>  			return 0;
>  
> -		if (*to_req == NULL) {
> -			struct drm_i915_gem_request *req;
> -
> -			req = i915_gem_request_alloc(to, NULL);
> -			if (IS_ERR(req))
> -				return PTR_ERR(req);
> -
> -			*to_req = req;
> -		}
> -
> -		trace_i915_gem_ring_sync_to(*to_req, from, from_req);
> -		ret = to->semaphore.sync_to(*to_req, from, seqno);
> +		trace_i915_gem_ring_sync_to(to, from);
> +		ret = to->engine->semaphore.sync_to(to, from->engine, seqno);
>  		if (ret)
>  			return ret;
>  
> @@ -2897,8 +2881,8 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>  		 * might have just caused seqno wrap under
>  		 * the radar.
>  		 */
> -		from->semaphore.sync_seqno[idx] =
> -			i915_gem_request_get_seqno(obj->last_read_req[from->id]);
> +		from->engine->semaphore.sync_seqno[idx] =
> +			i915_gem_request_get_seqno(obj->last_read_req[from->engine->id]);
>  	}
>  
>  	return 0;
> @@ -2908,17 +2892,12 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>   * i915_gem_object_sync - sync an object to a ring.
>   *
>   * @obj: object which may be in use on another ring.
> - * @to: ring we wish to use the object on. May be NULL.
> - * @to_req: request we wish to use the object for. See below.
> - *          This will be allocated and returned if a request is
> - *          required but not passed in.
> + * @to: request we are wishing to use
>   *
>   * This code is meant to abstract object synchronization with the GPU.
> - * Calling with NULL implies synchronizing the object with the CPU
> - * rather than a particular GPU ring. Conceptually we serialise writes
> - * between engines inside the GPU. We only allow one engine to write
> - * into a buffer at any time, but multiple readers. To ensure each has
> - * a coherent view of memory, we must:
> + * Conceptually we serialise writes between engines inside the GPU.
> + * We only allow one engine to write into a buffer at any time, but
> + * multiple readers. To ensure each has a coherent view of memory, we must:
>   *
>   * - If there is an outstanding write request to the object, the new
>   *   request must wait for it to complete (either CPU or in hw, requests
> @@ -2927,22 +2906,11 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>   * - If we are a write request (pending_write_domain is set), the new
>   *   request must wait for outstanding read requests to complete.
>   *
> - * For CPU synchronisation (NULL to) no request is required. For syncing with
> - * rings to_req must be non-NULL. However, a request does not have to be
> - * pre-allocated. If *to_req is NULL and sync commands will be emitted then a
> - * request will be allocated automatically and returned through *to_req. Note
> - * that it is not guaranteed that commands will be emitted (because the system
> - * might already be idle). Hence there is no need to create a request that
> - * might never have any work submitted. Note further that if a request is
> - * returned in *to_req, it is the responsibility of the caller to submit
> - * that request (after potentially adding more work to it).
> - *
>   * Returns 0 if successful, else propagates up the lower layer error.
>   */
>  int
>  i915_gem_object_sync(struct drm_i915_gem_object *obj,
> -		     struct intel_engine_cs *to,
> -		     struct drm_i915_gem_request **to_req)
> +		     struct drm_i915_gem_request *to)
>  {
>  	const bool readonly = obj->base.pending_write_domain == 0;
>  	struct drm_i915_gem_request *req[I915_NUM_ENGINES];
> @@ -2951,9 +2919,6 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
>  	if (!obj->active)
>  		return 0;
>  
> -	if (to == NULL)
> -		return i915_gem_object_wait_rendering(obj, readonly);
> -
>  	n = 0;
>  	if (readonly) {
>  		if (obj->last_write_req)
> @@ -2964,7 +2929,7 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
>  				req[n++] = obj->last_read_req[i];
>  	}
>  	for (i = 0; i < n; i++) {
> -		ret = __i915_gem_object_sync(obj, to, req[i], to_req);
> +		ret = __i915_gem_object_sync(obj, to, req[i]);
>  		if (ret)
>  			return ret;
>  	}
> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> index 048050176ff9..75d19932847d 100644
> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> @@ -981,7 +981,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
>  		struct drm_i915_gem_object *obj = vma->obj;
>  
>  		if (obj->active & other_rings) {
> -			ret = i915_gem_object_sync(obj, req->engine, &req);
> +			ret = i915_gem_object_sync(obj, req);
>  			if (ret)
>  				return ret;
>  		}
> @@ -1427,7 +1427,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>  {
>  	struct drm_i915_private *dev_priv = to_i915(dev);
>  	struct i915_ggtt *ggtt = &dev_priv->ggtt;
> -	struct drm_i915_gem_request *req = NULL;
>  	struct eb_vmas *eb;
>  	struct drm_i915_gem_object *batch_obj;
>  	struct drm_i915_gem_exec_object2 shadow_exec_entry;
> @@ -1615,13 +1614,13 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>  		params->batch_obj_vm_offset = i915_gem_obj_offset(batch_obj, vm);
>  
>  	/* Allocate a request for this batch buffer nice and early. */
> -	req = i915_gem_request_alloc(engine, ctx);
> -	if (IS_ERR(req)) {
> -		ret = PTR_ERR(req);
> +	params->request = i915_gem_request_alloc(engine, ctx);
> +	if (IS_ERR(params->request)) {
> +		ret = PTR_ERR(params->request);
>  		goto err_batch_unpin;
>  	}
>  
> -	ret = i915_gem_request_add_to_client(req, file);
> +	ret = i915_gem_request_add_to_client(params->request, file);
>  	if (ret)
>  		goto err_request;
>  
> @@ -1637,7 +1636,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>  	params->dispatch_flags          = dispatch_flags;
>  	params->batch_obj               = batch_obj;
>  	params->ctx                     = ctx;
> -	params->request                 = req;
>  
>  	ret = dev_priv->gt.execbuf_submit(params, args, &eb->vmas);
>  err_request:
> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 7b772d914e23..995ae99e9cf7 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -292,10 +292,21 @@ static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno)
>  	return 0;
>  }
>  
> -static inline int
> -__i915_gem_request_alloc(struct intel_engine_cs *engine,
> -			 struct i915_gem_context *ctx,
> -			 struct drm_i915_gem_request **req_out)
> +/**
> + * i915_gem_request_alloc - allocate a request structure
> + *
> + * @engine: engine that we wish to issue the request on.
> + * @ctx: context that the request will be associated with.
> + *       This can be NULL if the request is not directly related to
> + *       any specific user context, in which case this function will
> + *       choose an appropriate context to use.
> + *
> + * Returns a pointer to the allocated request if successful,
> + * or an error code if not.
> + */
> +struct drm_i915_gem_request *
> +i915_gem_request_alloc(struct intel_engine_cs *engine,
> +		       struct i915_gem_context *ctx)
>  {
>  	struct drm_i915_private *dev_priv = engine->i915;
>  	unsigned int reset_counter = i915_reset_counter(&dev_priv->gpu_error);
> @@ -303,18 +314,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>  	u32 seqno;
>  	int ret;
>  
> -	if (!req_out)
> -		return -EINVAL;
> -
> -	*req_out = NULL;
> -
>  	/* ABI: Before userspace accesses the GPU (e.g. execbuffer), report
>  	 * EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex
>  	 * and restart.
>  	 */
>  	ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible);
>  	if (ret)
> -		return ret;
> +		return ERR_PTR(ret);
>  
>  	/* Move the oldest request to the slab-cache (if not in use!) */
>  	req = list_first_entry_or_null(&engine->request_list,
> @@ -324,7 +330,7 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>  
>  	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
>  	if (!req)
> -		return -ENOMEM;
> +		return ERR_PTR(-ENOMEM);
>  
>  	ret = i915_gem_get_seqno(dev_priv, &seqno);
>  	if (ret)
> @@ -357,39 +363,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>  	if (ret)
>  		goto err_ctx;
>  
> -	*req_out = req;
> -	return 0;
> +	return req;
>  
>  err_ctx:
>  	i915_gem_context_put(ctx);
>  err:
>  	kmem_cache_free(dev_priv->requests, req);
> -	return ret;
> -}
> -
> -/**
> - * i915_gem_request_alloc - allocate a request structure
> - *
> - * @engine: engine that we wish to issue the request on.
> - * @ctx: context that the request will be associated with.
> - *       This can be NULL if the request is not directly related to
> - *       any specific user context, in which case this function will
> - *       choose an appropriate context to use.
> - *
> - * Returns a pointer to the allocated request if successful,
> - * or an error code if not.
> - */
> -struct drm_i915_gem_request *
> -i915_gem_request_alloc(struct intel_engine_cs *engine,
> -		       struct i915_gem_context *ctx)
> -{
> -	struct drm_i915_gem_request *req;
> -	int err;
> -
> -	if (!ctx)
> -		ctx = engine->i915->kernel_context;
> -	err = __i915_gem_request_alloc(engine, ctx, &req);
> -	return err ? ERR_PTR(err) : req;
> +	return ERR_PTR(ret);
>  }
>  
>  static void i915_gem_mark_busy(const struct intel_engine_cs *engine)
> diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
> index 007112d1e049..9e43c0aa6e3b 100644
> --- a/drivers/gpu/drm/i915/i915_trace.h
> +++ b/drivers/gpu/drm/i915/i915_trace.h
> @@ -449,10 +449,9 @@ TRACE_EVENT(i915_gem_evict_vm,
>  );
>  
>  TRACE_EVENT(i915_gem_ring_sync_to,
> -	    TP_PROTO(struct drm_i915_gem_request *to_req,
> -		     struct intel_engine_cs *from,
> -		     struct drm_i915_gem_request *req),
> -	    TP_ARGS(to_req, from, req),
> +	    TP_PROTO(struct drm_i915_gem_request *to,
> +		     struct drm_i915_gem_request *from),
> +	    TP_ARGS(to, from),
>  
>  	    TP_STRUCT__entry(
>  			     __field(u32, dev)
> @@ -463,9 +462,9 @@ TRACE_EVENT(i915_gem_ring_sync_to,
>  
>  	    TP_fast_assign(
>  			   __entry->dev = from->i915->drm.primary->index;
> -			   __entry->sync_from = from->id;
> -			   __entry->sync_to = to_req->engine->id;
> -			   __entry->seqno = req->fence.seqno;
> +			   __entry->sync_from = from->engine->id;
> +			   __entry->sync_to = to->engine->id;
> +			   __entry->seqno = from->fence.seqno;
>  			   ),
>  
>  	    TP_printk("dev=%u, sync-from=%u, sync-to=%u, seqno=%u",
> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> index bff172c45ff7..5d4420b67642 100644
> --- a/drivers/gpu/drm/i915/intel_display.c
> +++ b/drivers/gpu/drm/i915/intel_display.c
> @@ -11583,7 +11583,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>  	struct intel_flip_work *work;
>  	struct intel_engine_cs *engine;
>  	bool mmio_flip;
> -	struct drm_i915_gem_request *request = NULL;
> +	struct drm_i915_gem_request *request;
>  	int ret;
>  
>  	/*
> @@ -11690,22 +11690,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>  
>  	mmio_flip = use_mmio_flip(engine, obj);
>  
> -	/* When using CS flips, we want to emit semaphores between rings.
> -	 * However, when using mmio flips we will create a task to do the
> -	 * synchronisation, so all we want here is to pin the framebuffer
> -	 * into the display plane and skip any waits.
> -	 */
> -	if (!mmio_flip) {
> -		ret = i915_gem_object_sync(obj, engine, &request);
> -		if (!ret && !request) {
> -			request = i915_gem_request_alloc(engine, NULL);
> -			ret = PTR_ERR_OR_ZERO(request);
> -		}
> -
> -		if (ret)
> -			goto cleanup_pending;
> -	}
> -
>  	ret = intel_pin_and_fence_fb_obj(fb, primary->state->rotation);
>  	if (ret)
>  		goto cleanup_pending;
> @@ -11723,14 +11707,24 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>  
>  		schedule_work(&work->mmio_work);
>  	} else {
> -		i915_gem_request_assign(&work->flip_queued_req, request);
> +		request = i915_gem_request_alloc(engine, engine->last_context);
> +		if (IS_ERR(request)) {
> +			ret = PTR_ERR(request);
> +			goto cleanup_unpin;
> +		}
> +
> +		ret = i915_gem_object_sync(obj, request);
> +		if (ret)
> +			goto cleanup_request;
> +
>  		ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, request,
>  						   page_flip_flags);
>  		if (ret)
> -			goto cleanup_unpin;
> +			goto cleanup_request;
>  
>  		intel_mark_page_flip_active(intel_crtc, work);
>  
> +		work->flip_queued_req = i915_gem_request_get(request);
>  		i915_add_request_no_flush(request);
>  	}
>  
> @@ -11745,11 +11739,11 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>  
>  	return 0;
>  
> +cleanup_request:
> +	i915_add_request_no_flush(request);
>  cleanup_unpin:
>  	intel_unpin_fb_obj(fb, crtc->primary->state->rotation);
>  cleanup_pending:
> -	if (!IS_ERR_OR_NULL(request))
> -		i915_add_request_no_flush(request);
>  	atomic_dec(&intel_crtc->unpin_work_count);
>  	mutex_unlock(&dev->struct_mutex);
>  cleanup:
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index 6487794ac96c..a37af1533618 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -655,7 +655,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>  		struct drm_i915_gem_object *obj = vma->obj;
>  
>  		if (obj->active & other_rings) {
> -			ret = i915_gem_object_sync(obj, req->engine, &req);
> +			ret = i915_gem_object_sync(obj, req);
>  			if (ret)
>  				return ret;
>  		}
> diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
> index 8f1d4d9ef345..651efe4e468e 100644
> --- a/drivers/gpu/drm/i915/intel_overlay.c
> +++ b/drivers/gpu/drm/i915/intel_overlay.c
> @@ -229,11 +229,18 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
>  	return 0;
>  }
>  
> +static struct drm_i915_gem_request *alloc_request(struct intel_overlay *overlay)
> +{
> +	struct drm_i915_private *dev_priv = overlay->i915;
> +	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
> +
> +	return i915_gem_request_alloc(engine, dev_priv->kernel_context);
> +}
> +
>  /* overlay needs to be disable in OCMD reg */
>  static int intel_overlay_on(struct intel_overlay *overlay)
>  {
>  	struct drm_i915_private *dev_priv = overlay->i915;
> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>  	struct drm_i915_gem_request *req;
>  	struct intel_ring *ring;
>  	int ret;
> @@ -241,7 +248,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
>  	WARN_ON(overlay->active);
>  	WARN_ON(IS_I830(dev_priv) && !(dev_priv->quirks & QUIRK_PIPEA_FORCE));
>  
> -	req = i915_gem_request_alloc(engine, NULL);
> +	req = alloc_request(overlay);
>  	if (IS_ERR(req))
>  		return PTR_ERR(req);
>  
> @@ -268,7 +275,6 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
>  				  bool load_polyphase_filter)
>  {
>  	struct drm_i915_private *dev_priv = overlay->i915;
> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>  	struct drm_i915_gem_request *req;
>  	struct intel_ring *ring;
>  	u32 flip_addr = overlay->flip_addr;
> @@ -285,7 +291,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
>  	if (tmp & (1 << 17))
>  		DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp);
>  
> -	req = i915_gem_request_alloc(engine, NULL);
> +	req = alloc_request(overlay);
>  	if (IS_ERR(req))
>  		return PTR_ERR(req);
>  
> @@ -338,7 +344,6 @@ static void intel_overlay_off_tail(struct intel_overlay *overlay)
>  static int intel_overlay_off(struct intel_overlay *overlay)
>  {
>  	struct drm_i915_private *dev_priv = overlay->i915;
> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>  	struct drm_i915_gem_request *req;
>  	struct intel_ring *ring;
>  	u32 flip_addr = overlay->flip_addr;
> @@ -352,7 +357,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
>  	 * of the hw. Do it in both cases */
>  	flip_addr |= OFC_UPDATE;
>  
> -	req = i915_gem_request_alloc(engine, NULL);
> +	req = alloc_request(overlay);
>  	if (IS_ERR(req))
>  		return PTR_ERR(req);
>  
> @@ -412,7 +417,6 @@ static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay)
>  static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
>  {
>  	struct drm_i915_private *dev_priv = overlay->i915;
> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>  	int ret;
>  
>  	lockdep_assert_held(&dev_priv->drm.struct_mutex);
> @@ -428,7 +432,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
>  		struct drm_i915_gem_request *req;
>  		struct intel_ring *ring;
>  
> -		req = i915_gem_request_alloc(engine, NULL);
> +		req = alloc_request(overlay);
>  		if (IS_ERR(req))
>  			return PTR_ERR(req);
>  
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Chris Wilson]

On Mon, Jul 25, 2016 at 12:14:44PM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > Space for flushing the GPU cache prior to completing the request is
> > preallocated and so cannot fail.
> 
> Patch title and commit message have some disconnect. Could you explain
> in a bit more detail what made gpu_caches_dirty obsolete?

Introduction of requests. gpu_caches_dirty tracked the GPU cache state
for the outstanding_lazy_seqno, long obsolete, since we only submit
complete requests. (Previously we could cancel the breadcrumb due to an
interrupt without having flushed the caches and so before the next batch
we had to emit a cache flush.)
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Ro.CI.BAT: success for series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev2)

[Patchwork]

== Series Details ==

Series: series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev2)
URL   : https://patchwork.freedesktop.org/series/10230/
State : success

== Summary ==

Series 10230v2 Series without cover letter
http://patchwork.freedesktop.org/api/1.0/series/10230/revisions/2/mbox


ro-bdw-i5-5250u  total:244  pass:219  dwarn:4   dfail:0   fail:8   skip:13 
ro-bdw-i7-5557U  total:244  pass:221  dwarn:1   dfail:0   fail:8   skip:14 
ro-bdw-i7-5600u  total:244  pass:204  dwarn:0   dfail:0   fail:8   skip:32 
ro-bsw-n3050     total:218  pass:173  dwarn:0   dfail:0   fail:2   skip:42 
ro-byt-n2820     total:244  pass:197  dwarn:0   dfail:0   fail:9   skip:38 
ro-hsw-i3-4010u  total:244  pass:212  dwarn:0   dfail:0   fail:8   skip:24 
ro-hsw-i7-4770r  total:244  pass:212  dwarn:0   dfail:0   fail:8   skip:24 
ro-ilk-i7-620lm  total:244  pass:172  dwarn:0   dfail:0   fail:9   skip:63 
ro-ilk1-i5-650   total:239  pass:172  dwarn:0   dfail:0   fail:9   skip:58 
ro-ivb-i7-3770   total:244  pass:203  dwarn:0   dfail:0   fail:8   skip:33 
ro-skl3-i5-6260u total:244  pass:224  dwarn:0   dfail:0   fail:8   skip:12 
ro-snb-i7-2620M  total:244  pass:193  dwarn:0   dfail:0   fail:9   skip:42 

Results at /archive/results/CI_IGT_test/RO_Patchwork_1597/

5c9e3d9 drm-intel-nightly: 2016y-07m-25d-06h-32m-37s UTC integration manifest
bedeecf drm/i915: Rename engine->semaphore.sync_to, engine->sempahore.signal locals
7e586ab drm/i915: Simplify calling engine->sync_to
542fe17 drm/i915: Unify legacy/execlists submit_execbuf callbacks
ff10e7e drm/i915: Refactor golden render state emission to unconfuse gcc
32095db drm/i915: Remove duplicate golden render state init from execlists
2a3e2d7 drm/i915/ringbuffer: Specialise SNB+ request emission for semaphores
a622ad64b drm/i915: Reuse legacy breadcrumbs + tail emission
4446715 drm/i915: Stop passing caller's num_dwords to engine->semaphore.signal()
50f4681 drm/i915/lrc: Update function names to match request flow
39e9cef drm/i915: Unify request submission
ae05ba0 drm/i915: Convert engine->write_tail to operate on a request
5495322 drm/i915: Remove intel_ring_get_tail()
14e06f3 drm/i915: Unify legacy/execlists emission of MI_BATCHBUFFER_START
e337284 drm/i915: Simplify request_alloc by returning the allocated request
47701d5 drm/i915: Remove obsolete engine->gpu_caches_dirty
c1cdbc6 drm/i915: Rename intel_pin_and_map_ring()
a3b5cc8 drm/i915: Rename residual ringbuf parameters
b296387 drm/i915: Rename struct intel_ringbuffer to struct intel_ring
9bdf7e2 drm/i915: Rename intel_context[engine].ringbuf
3ae54a2e drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs
b4b003c drm/i915: Rename request->ringbuf to request->ring
d40ec77 drm/i915: Unify intel_logical_ring_emit and intel_ring_emit
2b302ad drm/i915: Update a couple of hangcheck comments to talk about engines
dc1807f drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c
38001ff drm/i915: Avoid using intel_engine_cs *ring for GPU error capture
f68488d drm/i915: Use engine to refer to the user's BSD intel_engine_cs
ce8c1ab drm/i915/cmdparser: Remove stray intel_engine_cs *ring
5596a46 drm/i915: Only drop the batch-pool's object reference
0ab690f drm/i915: Only clear the client pointer when tearing down the file
7c850c6 drm/i915: Prefer list_first_entry_or_null
baf2c02 drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 20/31] drm/i915: Remove intel_ring_get_tail()

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Joonas doesn't like the tiny function, especially if I go around making
> it more complicated and using it elsewhere. To remove that temptation,
> remove the function!
> 

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_gem_request.c | 8 ++++----
>  drivers/gpu/drm/i915/intel_ringbuffer.h | 5 -----
>  2 files changed, 4 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 995ae99e9cf7..1c185e293bf0 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -419,7 +419,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  	 * should already have been reserved in the ring buffer. Let the ring
>  	 * know that it is time to use that space up.
>  	 */
> -	request_start = intel_ring_get_tail(ring);
> +	request_start = ring->tail;
>  	reserved_tail = request->reserved_space;
>  	request->reserved_space = 0;
>  
> @@ -465,19 +465,19 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  	 * GPU processing the request, we never over-estimate the
>  	 * position of the head.
>  	 */
> -	request->postfix = intel_ring_get_tail(ring);
> +	request->postfix = ring->tail;
>  
>  	if (i915.enable_execlists) {
>  		ret = engine->emit_request(request);
>  	} else {
>  		ret = engine->add_request(request);
>  
> -		request->tail = intel_ring_get_tail(ring);
> +		request->tail = ring->tail;
>  	}
>  	/* Not allowed to fail! */
>  	WARN(ret, "emit|add_request failed: %d!\n", ret);
>  	/* Sanity check that the reserved size was large enough. */
> -	ret = intel_ring_get_tail(ring) - request_start;
> +	ret = ring->tail - request_start;
>  	if (ret < 0)
>  		ret += ring->size;
>  	WARN_ONCE(ret > reserved_tail,
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index cbd46d5e8d6b..13b816f2d264 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -488,11 +488,6 @@ static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine)
>  
>  int init_workarounds_ring(struct intel_engine_cs *engine);
>  
> -static inline u32 intel_ring_get_tail(struct intel_ring *ring)
> -{
> -	return ring->tail;
> -}
> -
>  /*
>   * Arbitrary size for largest possible 'add request' sequence. The code paths
>   * are complex and variable. Empirical measurement shows that the worst case
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 22/31] drm/i915: Unify request submission

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Move request submission from emit_request into its own common vfunc
> from i915_add_request().
> 
> v2: Convert I915_DISPATCH_flags to BIT(x) whilst passing
> v3: Rename a few functions to match.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/i915_gem_request.c    |  8 +++-----
>  drivers/gpu/drm/i915/i915_guc_submission.c |  9 ++++++---
>  drivers/gpu/drm/i915/intel_guc.h           |  1 -
>  drivers/gpu/drm/i915/intel_lrc.c           | 18 +++++++-----------
>  drivers/gpu/drm/i915/intel_ringbuffer.c    | 23 +++++++++--------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h    | 23 +++++++++++------------
>  6 files changed, 36 insertions(+), 46 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 8814e9c5266b..f4bf9f669eed 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -467,12 +467,9 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  	 */
>  	request->postfix = ring->tail;
>  
> -	if (i915.enable_execlists)
> -		ret = engine->emit_request(request);
> -	else
> -		ret = engine->add_request(request);
>  	/* Not allowed to fail! */
> -	WARN(ret, "emit|add_request failed: %d!\n", ret);
> +	ret = engine->emit_request(request);
> +	WARN(ret, "(%s)->emit_request failed: %d!\n", engine->name, ret);
>  
>  	/* Sanity check that the reserved size was large enough. */
>  	ret = ring->tail - request_start;
> @@ -484,6 +481,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  		  reserved_tail, ret);
>  
>  	i915_gem_mark_busy(engine);
> +	engine->submit_request(request);
>  }
>  
>  static unsigned long local_clock_us(unsigned int *cpu)
> diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
> index eccd34832fe6..32d0e1890950 100644
> --- a/drivers/gpu/drm/i915/i915_guc_submission.c
> +++ b/drivers/gpu/drm/i915/i915_guc_submission.c
> @@ -585,7 +585,7 @@ static int guc_ring_doorbell(struct i915_guc_client *gc)
>   * The only error here arises if the doorbell hardware isn't functioning
>   * as expected, which really shouln't happen.
>   */
> -int i915_guc_submit(struct drm_i915_gem_request *rq)
> +static void i915_guc_submit(struct drm_i915_gem_request *rq)
>  {
>  	unsigned int engine_id = rq->engine->id;
>  	struct intel_guc *guc = &rq->i915->guc;
> @@ -602,8 +602,6 @@ int i915_guc_submit(struct drm_i915_gem_request *rq)
>  
>  	guc->submissions[engine_id] += 1;
>  	guc->last_seqno[engine_id] = rq->fence.seqno;
> -
> -	return b_ret;
>  }
>  
>  /*
> @@ -992,6 +990,7 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
>  {
>  	struct intel_guc *guc = &dev_priv->guc;
>  	struct i915_guc_client *client;
> +	struct intel_engine_cs *engine;
>  
>  	/* client for execbuf submission */
>  	client = guc_client_alloc(dev_priv,
> @@ -1006,6 +1005,10 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv)
>  	host2guc_sample_forcewake(guc, client);
>  	guc_init_doorbell_hw(guc);
>  
> +	/* Take over from manual control of ELSP (execlists) */
> +	for_each_engine(engine, dev_priv)
> +		engine->submit_request = i915_guc_submit;
> +
>  	return 0;
>  }
>  
> diff --git a/drivers/gpu/drm/i915/intel_guc.h b/drivers/gpu/drm/i915/intel_guc.h
> index 3e3e743740c0..623cf26cd784 100644
> --- a/drivers/gpu/drm/i915/intel_guc.h
> +++ b/drivers/gpu/drm/i915/intel_guc.h
> @@ -160,7 +160,6 @@ extern int intel_guc_resume(struct drm_device *dev);
>  int i915_guc_submission_init(struct drm_i915_private *dev_priv);
>  int i915_guc_submission_enable(struct drm_i915_private *dev_priv);
>  int i915_guc_wq_check_space(struct drm_i915_gem_request *rq);
> -int i915_guc_submit(struct drm_i915_gem_request *rq);
>  void i915_guc_submission_disable(struct drm_i915_private *dev_priv);
>  void i915_guc_submission_fini(struct drm_i915_private *dev_priv);
>  
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index 250edb2bcef7..a9ca31c113c3 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -738,7 +738,7 @@ err_unpin:
>  }
>  
>  /*
> - * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
> + * intel_logical_ring_advance() - advance the tail and prepare for submission
>   * @request: Request to advance the logical ringbuffer of.
>   *
>   * The tail is updated in our logical ringbuffer struct, not in the actual context. What
> @@ -747,7 +747,7 @@ err_unpin:
>   * point, the tail *inside* the context is updated and the ELSP written to.
>   */
>  static int
> -intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
> +intel_logical_ring_advance(struct drm_i915_gem_request *request)
>  {
>  	struct intel_ring *ring = request->ring;
>  	struct intel_engine_cs *engine = request->engine;
> @@ -773,12 +773,6 @@ intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
>  	 */
>  	request->previous_context = engine->last_context;
>  	engine->last_context = request->ctx;
> -
> -	if (i915.enable_guc_submission)
> -		i915_guc_submit(request);
> -	else
> -		execlists_context_queue(request);
> -
>  	return 0;
>  }
>  
> @@ -1775,7 +1769,7 @@ static int gen8_emit_request(struct drm_i915_gem_request *request)
>  	intel_ring_emit(ring, request->fence.seqno);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
>  	intel_ring_emit(ring, MI_NOOP);
> -	return intel_logical_ring_advance_and_submit(request);
> +	return intel_logical_ring_advance(request);
>  }
>  
>  static int gen8_emit_request_render(struct drm_i915_gem_request *request)
> @@ -1806,7 +1800,7 @@ static int gen8_emit_request_render(struct drm_i915_gem_request *request)
>  	intel_ring_emit(ring, 0);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
>  	intel_ring_emit(ring, MI_NOOP);
> -	return intel_logical_ring_advance_and_submit(request);
> +	return intel_logical_ring_advance(request);
>  }
>  
>  static int intel_lr_context_render_state_init(struct drm_i915_gem_request *req)
> @@ -1912,8 +1906,10 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
>  {
>  	/* Default vfuncs which can be overriden by each engine. */
>  	engine->init_hw = gen8_init_common_ring;
> -	engine->emit_request = gen8_emit_request;
>  	engine->emit_flush = gen8_emit_flush;
> +	engine->emit_request = gen8_emit_request;
> +	engine->submit_request = execlists_context_queue;
> +
>  	engine->irq_enable = gen8_logical_ring_enable_irq;
>  	engine->irq_disable = gen8_logical_ring_disable_irq;
>  	engine->emit_bb_start = gen8_emit_bb_start;
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 3e1049c972e0..2fa7db5331c3 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -1441,15 +1441,14 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
>  }
>  
>  /**
> - * gen6_add_request - Update the semaphore mailbox registers
> + * gen6_emit_request - Update the semaphore mailbox registers
>   *
>   * @request - request to write to the ring
>   *
>   * Update the mailbox registers in the *other* rings with the current seqno.
>   * This acts like a signal in the canonical semaphore.
>   */
> -static int
> -gen6_add_request(struct drm_i915_gem_request *req)
> +static int gen6_emit_request(struct drm_i915_gem_request *req)
>  {
>  	struct intel_engine_cs *engine = req->engine;
>  	struct intel_ring *ring = req->ring;
> @@ -1470,13 +1469,11 @@ gen6_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_advance(ring);
>  
>  	req->tail = ring->tail;
> -	engine->submit_request(req);
>  
>  	return 0;
>  }
>  
> -static int
> -gen8_render_add_request(struct drm_i915_gem_request *req)
> +static int gen8_render_emit_request(struct drm_i915_gem_request *req)
>  {
>  	struct intel_engine_cs *engine = req->engine;
>  	struct intel_ring *ring = req->ring;
> @@ -1500,9 +1497,9 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_emit(ring, 0);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
>  	intel_ring_emit(ring, MI_NOOP);
> +	intel_ring_advance(ring);
>  
>  	req->tail = ring->tail;
> -	engine->submit_request(req);
>  
>  	return 0;
>  }
> @@ -1707,8 +1704,7 @@ bsd_ring_flush(struct drm_i915_gem_request *req,
>  	return 0;
>  }
>  
> -static int
> -i9xx_add_request(struct drm_i915_gem_request *req)
> +static int i9xx_emit_request(struct drm_i915_gem_request *req)
>  {
>  	struct intel_ring *ring = req->ring;
>  	int ret;
> @@ -1724,7 +1720,6 @@ i9xx_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_advance(ring);
>  
>  	req->tail = ring->tail;
> -	req->engine->submit_request(req);
>  
>  	return 0;
>  }
> @@ -2829,11 +2824,11 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
>  				      struct intel_engine_cs *engine)
>  {
>  	engine->init_hw = init_ring_common;
> -	engine->submit_request = i9xx_submit_request;
>  
> -	engine->add_request = i9xx_add_request;
> +	engine->emit_request = i9xx_emit_request;
>  	if (INTEL_GEN(dev_priv) >= 6)
> -		engine->add_request = gen6_add_request;
> +		engine->emit_request = gen6_emit_request;
> +	engine->submit_request = i9xx_submit_request;
>  
>  	if (INTEL_GEN(dev_priv) >= 8)
>  		engine->emit_bb_start = gen8_emit_bb_start;
> @@ -2862,7 +2857,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine)
>  
>  	if (INTEL_GEN(dev_priv) >= 8) {
>  		engine->init_context = intel_rcs_ctx_init;
> -		engine->add_request = gen8_render_add_request;
> +		engine->emit_request = gen8_render_emit_request;
>  		engine->emit_flush = gen8_render_ring_flush;
>  		if (i915.semaphores)
>  			engine->semaphore.signal = gen8_rcs_signal;
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 5428a3c288d5..fdf085495e3a 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -204,7 +204,17 @@ struct intel_engine_cs {
>  
>  	int		(*init_context)(struct drm_i915_gem_request *req);
>  
> -	int		(*add_request)(struct drm_i915_gem_request *req);
> +	int		(*emit_flush)(struct drm_i915_gem_request *request,
> +				      u32 invalidate_domains,
> +				      u32 flush_domains);
> +	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
> +					 u64 offset, u32 length,
> +					 unsigned int dispatch_flags);
> +#define I915_DISPATCH_SECURE BIT(0)
> +#define I915_DISPATCH_PINNED BIT(1)
> +#define I915_DISPATCH_RS     BIT(2)
> +	int		(*emit_request)(struct drm_i915_gem_request *req);
> +	void		(*submit_request)(struct drm_i915_gem_request *req);
>  	/* Some chipsets are not quite as coherent as advertised and need
>  	 * an expensive kick to force a true read of the up-to-date seqno.
>  	 * However, the up-to-date seqno is not always required and the last
> @@ -282,17 +292,6 @@ struct intel_engine_cs {
>  	unsigned int idle_lite_restore_wa;
>  	bool disable_lite_restore_wa;
>  	u32 ctx_desc_template;
> -	int		(*emit_request)(struct drm_i915_gem_request *request);
> -	int		(*emit_flush)(struct drm_i915_gem_request *request,
> -				      u32 invalidate_domains,
> -				      u32 flush_domains);
> -	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
> -					 u64 offset, u32 length,
> -					 unsigned int dispatch_flags);
> -#define I915_DISPATCH_SECURE 0x1
> -#define I915_DISPATCH_PINNED 0x2
> -#define I915_DISPATCH_RS     0x4
> -	void		(*submit_request)(struct drm_i915_gem_request *req);
>  
>  	/**
>  	 * List of objects currently involved in rendering from the
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 23/31] drm/i915/lrc: Update function names to match request flow

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> With adding engine->submit_request, we now have a bunch of functions
> with similar names used at different stages of the execlist submission.
> Try a different coat of paint, to hopefully reduce confusion between the
> requests, intel_engine_cs and the actual execlists submision process.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

You forgot to pick my R-b from previous series, to make it easier for
Patchwork, here it is again;

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_lrc.c | 16 ++++++++--------
>  1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index a9ca31c113c3..149a0dc7aeed 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -384,8 +384,8 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
>  		execlists_update_context_pdps(ppgtt, reg_state);
>  }
>  
> -static void execlists_submit_requests(struct drm_i915_gem_request *rq0,
> -				      struct drm_i915_gem_request *rq1)
> +static void execlists_elsp_submit_contexts(struct drm_i915_gem_request *rq0,
> +					   struct drm_i915_gem_request *rq1)
>  {
>  	struct drm_i915_private *dev_priv = rq0->i915;
>  	unsigned int fw_domains = rq0->engine->fw_domains;
> @@ -418,7 +418,7 @@ static inline void execlists_context_status_change(
>  	atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq);
>  }
>  
> -static void execlists_context_unqueue(struct intel_engine_cs *engine)
> +static void execlists_unqueue(struct intel_engine_cs *engine)
>  {
>  	struct drm_i915_gem_request *req0 = NULL, *req1 = NULL;
>  	struct drm_i915_gem_request *cursor, *tmp;
> @@ -486,7 +486,7 @@ static void execlists_context_unqueue(struct intel_engine_cs *engine)
>  		req0->tail &= req0->ring->size - 1;
>  	}
>  
> -	execlists_submit_requests(req0, req1);
> +	execlists_elsp_submit_contexts(req0, req1);
>  }
>  
>  static unsigned int
> @@ -597,7 +597,7 @@ static void intel_lrc_irq_handler(unsigned long data)
>  	if (submit_contexts) {
>  		if (!engine->disable_lite_restore_wa ||
>  		    (csb[i][0] & GEN8_CTX_STATUS_ACTIVE_IDLE))
> -			execlists_context_unqueue(engine);
> +			execlists_unqueue(engine);
>  	}
>  
>  	spin_unlock(&engine->execlist_lock);
> @@ -606,7 +606,7 @@ static void intel_lrc_irq_handler(unsigned long data)
>  		DRM_ERROR("More than two context complete events?\n");
>  }
>  
> -static void execlists_context_queue(struct drm_i915_gem_request *request)
> +static void execlists_submit_request(struct drm_i915_gem_request *request)
>  {
>  	struct intel_engine_cs *engine = request->engine;
>  	struct drm_i915_gem_request *cursor;
> @@ -637,7 +637,7 @@ static void execlists_context_queue(struct drm_i915_gem_request *request)
>  	list_add_tail(&request->execlist_link, &engine->execlist_queue);
>  	request->ctx_hw_id = request->ctx->hw_id;
>  	if (num_elements == 0)
> -		execlists_context_unqueue(engine);
> +		execlists_unqueue(engine);
>  
>  	spin_unlock_bh(&engine->execlist_lock);
>  }
> @@ -1908,7 +1908,7 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
>  	engine->init_hw = gen8_init_common_ring;
>  	engine->emit_flush = gen8_emit_flush;
>  	engine->emit_request = gen8_emit_request;
> -	engine->submit_request = execlists_context_queue;
> +	engine->submit_request = execlists_submit_request;
>  
>  	engine->irq_enable = gen8_logical_ring_enable_irq;
>  	engine->irq_disable = gen8_logical_ring_disable_irq;
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 24/31] drm/i915: Stop passing caller's num_dwords to engine->semaphore.signal()

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Rather than pass in the num_dwords that the caller wishes to use after
> the signal command packet, split the breadcrumb emission into two phases
> and have both the signal and breadcrumb individiually acquire space on
> the ring. This makes the interface simpler for the reader, and will
> simplify for patches.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 51 ++++++++++++++-------------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h |  4 +--
>  2 files changed, 23 insertions(+), 32 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 2fa7db5331c3..68df689a9d1e 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -1322,10 +1322,8 @@ static void render_ring_cleanup(struct intel_engine_cs *engine)
>  	intel_fini_pipe_control(engine);
>  }
>  
> -static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
> -			   unsigned int num_dwords)
> +static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req)
>  {
> -#define MBOX_UPDATE_DWORDS 8
>  	struct intel_ring *signaller = signaller_req->ring;
>  	struct drm_i915_private *dev_priv = signaller_req->i915;
>  	struct intel_engine_cs *waiter;
> @@ -1333,10 +1331,7 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
>  	int ret, num_rings;
>  
>  	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
> -	num_dwords += (num_rings-1) * MBOX_UPDATE_DWORDS;
> -#undef MBOX_UPDATE_DWORDS
> -
> -	ret = intel_ring_begin(signaller_req, num_dwords);
> +	ret = intel_ring_begin(signaller_req, (num_rings-1) * 8);
>  	if (ret)
>  		return ret;
>  
> @@ -1360,14 +1355,13 @@ static int gen8_rcs_signal(struct drm_i915_gem_request *signaller_req,
>  				MI_SEMAPHORE_TARGET(waiter->hw_id));
>  		intel_ring_emit(signaller, 0);
>  	}
> +	intel_ring_advance(signaller);
>  
>  	return 0;
>  }
>  
> -static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
> -			   unsigned int num_dwords)
> +static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req)
>  {
> -#define MBOX_UPDATE_DWORDS 6
>  	struct intel_ring *signaller = signaller_req->ring;
>  	struct drm_i915_private *dev_priv = signaller_req->i915;
>  	struct intel_engine_cs *waiter;
> @@ -1375,10 +1369,7 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
>  	int ret, num_rings;
>  
>  	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
> -	num_dwords += (num_rings-1) * MBOX_UPDATE_DWORDS;
> -#undef MBOX_UPDATE_DWORDS
> -
> -	ret = intel_ring_begin(signaller_req, num_dwords);
> +	ret = intel_ring_begin(signaller_req, (num_rings-1) * 6);
>  	if (ret)
>  		return ret;
>  
> @@ -1400,12 +1391,12 @@ static int gen8_xcs_signal(struct drm_i915_gem_request *signaller_req,
>  				MI_SEMAPHORE_TARGET(waiter->hw_id));
>  		intel_ring_emit(signaller, 0);
>  	}
> +	intel_ring_advance(signaller);
>  
>  	return 0;
>  }
>  
> -static int gen6_signal(struct drm_i915_gem_request *signaller_req,
> -		       unsigned int num_dwords)
> +static int gen6_signal(struct drm_i915_gem_request *signaller_req)
>  {
>  	struct intel_ring *signaller = signaller_req->ring;
>  	struct drm_i915_private *dev_priv = signaller_req->i915;
> @@ -1413,12 +1404,8 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
>  	enum intel_engine_id id;
>  	int ret, num_rings;
>  
> -#define MBOX_UPDATE_DWORDS 3
>  	num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
> -	num_dwords += round_up((num_rings-1) * MBOX_UPDATE_DWORDS, 2);
> -#undef MBOX_UPDATE_DWORDS
> -
> -	ret = intel_ring_begin(signaller_req, num_dwords);
> +	ret = intel_ring_begin(signaller_req, round_up((num_rings-1) * 3, 2));
>  	if (ret)
>  		return ret;
>  
> @@ -1436,6 +1423,7 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req,
>  	/* If num_dwords was rounded, make sure the tail pointer is correct */
>  	if (num_rings % 2 == 0)
>  		intel_ring_emit(signaller, MI_NOOP);
> +	intel_ring_advance(signaller);
>  
>  	return 0;
>  }
> @@ -1454,11 +1442,13 @@ static int gen6_emit_request(struct drm_i915_gem_request *req)
>  	struct intel_ring *ring = req->ring;
>  	int ret;
>  
> -	if (engine->semaphore.signal)
> -		ret = engine->semaphore.signal(req, 4);
> -	else
> -		ret = intel_ring_begin(req, 4);
> +	if (engine->semaphore.signal) {
> +		ret = engine->semaphore.signal(req);
> +		if (ret)
> +			return ret;
> +	}
>  
> +	ret = intel_ring_begin(req, 4);
>  	if (ret)
>  		return ret;
>  
> @@ -1479,10 +1469,13 @@ static int gen8_render_emit_request(struct drm_i915_gem_request *req)
>  	struct intel_ring *ring = req->ring;
>  	int ret;
>  
> -	if (engine->semaphore.signal)
> -		ret = engine->semaphore.signal(req, 8);
> -	else
> -		ret = intel_ring_begin(req, 8);
> +	if (engine->semaphore.signal) {
> +		ret = engine->semaphore.signal(req);
> +		if (ret)
> +			return ret;
> +	}
> +
> +	ret = intel_ring_begin(req, 8);
>  	if (ret)
>  		return ret;
>  
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index fdf085495e3a..a137fd218c83 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -278,9 +278,7 @@ struct intel_engine_cs {
>  		int	(*sync_to)(struct drm_i915_gem_request *to_req,
>  				   struct intel_engine_cs *from,
>  				   u32 seqno);
> -		int	(*signal)(struct drm_i915_gem_request *signaller_req,
> -				  /* num_dwords needed by caller */
> -				  unsigned int num_dwords);
> +		int	(*signal)(struct drm_i915_gem_request *signaller_req);
>  	} semaphore;
>  
>  	/* Execlists */
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 26/31] drm/i915/ringbuffer: Specialise SNB+ request emission for semaphores

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> As gen6_emit_request() only differs from i9xx_emit_request() when
> semaphores are enabled, only use the specialised vfunc in that scenario.
> 
> v2: Reorder semaphore init so as to keep engine->emit_request default
> vfunc selection compact.

With this change,

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 28/31] drm/i915: Refactor golden render state emission to unconfuse gcc

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> +	so.obj = i915_gem_object_create(&req->i915->drm, 4096);
> +	if (IS_ERR(so.obj))
> +		return PTR_ERR(so.obj);

I remember earlier discussion about having GPU_PAGE_SIZE, did that ever
go anywhere?

Nor related to this patch specifically, so;

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 02/31] drm/i915: Prefer list_first_entry_or_null

[Joonas Lahtinen]

On ma, 2016-07-25 at 09:03 +0100, Chris Wilson wrote:
> On Mon, Jul 25, 2016 at 10:55:24AM +0300, Joonas Lahtinen wrote:
> > 
> > On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > > 
> > > list_first_entry_or_null() can generate better code than using
> > > if (!list_empty()) {ptr = list_first_entry()) ..., so put it to use.
> > > 
> > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > It looks cleaner, not so sure of the code impact.
> After the patch I sent to linux-kernel@ to improve code gneration for
> list_first_entry_or_null() is applied we save a few bytes on each
> invocation.

Ok, I've not spotted those, but sounds good.

Regards, Joonas
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 04/31] drm/i915: Only drop the batch-pool's object reference

[Joonas Lahtinen]

On ma, 2016-07-25 at 09:44 +0100, Chris Wilson wrote:
> On Mon, Jul 25, 2016 at 11:38:07AM +0300, Joonas Lahtinen wrote:
> > 
> > On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> > > 
> > > The obj->batch_pool_link is only inspected when traversing the batch
> > > pool list and when on the batch pool list the object is referenced. Thus
> > > when freeing the batch pool list, we only need to unreference the object
> > > and do not have to worry about the obj->batch_pool_link.
> > > 
> > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > ---
> > >  drivers/gpu/drm/i915/i915_gem_batch_pool.c | 13 ++++++-------
> > >  1 file changed, 6 insertions(+), 7 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > > index 3507b2753fd3..825981b5aa40 100644
> > > --- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > > +++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
> > > @@ -68,15 +68,14 @@ void i915_gem_batch_pool_fini(struct i915_gem_batch_pool *pool)
> > >  	WARN_ON(!mutex_is_locked(&pool->dev->struct_mutex));
> > >  
> > >  	for (n = 0; n < ARRAY_SIZE(pool->cache_list); n++) {
> > > -		while (!list_empty(&pool->cache_list[n])) {
> > > -			struct drm_i915_gem_object *obj =
> > > -				list_first_entry(&pool->cache_list[n],
> > > -						 struct drm_i915_gem_object,
> > > -						 batch_pool_link);
> > > +		struct drm_i915_gem_object *obj, *next;
> > >  
> > > -			list_del(&obj->batch_pool_link);
> > > +		list_for_each_entry_safe(obj, next,
> > Why _safe? i915_gem_free_object does not touch the batch_pool_link.

Yeah, that'd be true.

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> Not directly, but it does free the memory including the batch_pool_link. :)
> -Chris
> 
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH v2] drm/i915/cmdparser: Remove stray intel_engine_cs *ring

[Joonas Lahtinen]

On ma, 2016-07-25 at 10:06 +0100, Chris Wilson wrote:
> When we refer to intel_engine_cs, we want to use engine so as not to
> confuse ourselves about ringbuffers.
> 
> v2: Rename all the functions as well, as well as a few more stray comments.
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Link: http://patchwork.freedesktop.org/patch/msgid/1469432687-22756-6-git-send-email-chris@chris-wilson.co.uk
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/i915_cmd_parser.c     | 72 +++++++++++++++---------------
>  drivers/gpu/drm/i915/i915_drv.h            | 23 +++++-----
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c | 14 +++---
>  drivers/gpu/drm/i915/intel_engine_cs.c     |  2 +-
>  drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
>  drivers/gpu/drm/i915/intel_ringbuffer.c    |  2 +-
>  drivers/gpu/drm/i915/intel_ringbuffer.h    | 10 ++---
>  7 files changed, 64 insertions(+), 61 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
> index b0fd6a7..8db144b 100644
> --- a/drivers/gpu/drm/i915/i915_cmd_parser.c
> +++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
> @@ -62,23 +62,23 @@
>   * The parser always rejects such commands.
>   *
>   * The majority of the problematic commands fall in the MI_* range, with only a
> - * few specific commands on each ring (e.g. PIPE_CONTROL and MI_FLUSH_DW).
> + * few specific commands on each engine (e.g. PIPE_CONTROL and MI_FLUSH_DW).
>   *
>   * Implementation:
> - * Each ring maintains tables of commands and registers which the parser uses in
> - * scanning batch buffers submitted to that ring.
> + * Each engine maintains tables of commands and registers which the parser
> + * uses in scanning batch buffers submitted to that engine.
>   *
>   * Since the set of commands that the parser must check for is significantly
>   * smaller than the number of commands supported, the parser tables contain only
>   * those commands required by the parser. This generally works because command
>   * opcode ranges have standard command length encodings. So for commands that
>   * the parser does not need to check, it can easily skip them. This is
> - * implemented via a per-ring length decoding vfunc.
> + * implemented via a per-engine length decoding vfunc.
>   *
>   * Unfortunately, there are a number of commands that do not follow the standard
>   * length encoding for their opcode range, primarily amongst the MI_* commands.
>   * To handle this, the parser provides a way to define explicit "skip" entries
> - * in the per-ring command tables.
> + * in the per-engine command tables.
>   *
>   * Other command table entries map fairly directly to high level categories
>   * mentioned above: rejected, master-only, register whitelist. The parser
> @@ -603,7 +603,7 @@ static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
>  	return 0;
>  }
>  
> -static bool validate_cmds_sorted(struct intel_engine_cs *engine,
> +static bool validate_cmds_sorted(const struct intel_engine_cs *engine,

This otherwise unrelated change is a dependency from a signature change
below.

>  				 const struct drm_i915_cmd_table *cmd_tables,
>  				 int cmd_table_count)
>  {
> @@ -624,8 +624,9 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
>  			u32 curr = desc->cmd.value & desc->cmd.mask;
>  
>  			if (curr < previous) {
> -				DRM_ERROR("CMD: table not sorted ring=%d table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
> -					  engine->id, i, j, curr, previous);
> +				DRM_ERROR("CMD: %s [%d] command table not sorted: table=%d entry=%d cmd=0x%08X prev=0x%08X\n",

Cut this line here like;

DRM_ERROR("CMD: %s [%d] command table not sorted: "
	  "table=%d entry=%d cmd=0x%08X prev=0x%08X\n"

> +					  engine->name, engine->id,
> +					  i, j, curr, previous);

Then this cut makes more sense, too.

Apart from that, looks fairly mechanical;

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH v2] drm/i915/cmdparser: Remove stray intel_engine_cs *ring

[Chris Wilson]

On Mon, Jul 25, 2016 at 02:01:05PM +0300, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 10:06 +0100, Chris Wilson wrote:
> > When we refer to intel_engine_cs, we want to use engine so as not to
> > confuse ourselves about ringbuffers.
> > 
> > v2: Rename all the functions as well, as well as a few more stray comments.
> > 
> > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > Link: http://patchwork.freedesktop.org/patch/msgid/1469432687-22756-6-git-send-email-chris@chris-wilson.co.uk
> > Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> > ---
> >  drivers/gpu/drm/i915/i915_cmd_parser.c     | 72 +++++++++++++++---------------
> >  drivers/gpu/drm/i915/i915_drv.h            | 23 +++++-----
> >  drivers/gpu/drm/i915/i915_gem_execbuffer.c | 14 +++---
> >  drivers/gpu/drm/i915/intel_engine_cs.c     |  2 +-
> >  drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
> >  drivers/gpu/drm/i915/intel_ringbuffer.c    |  2 +-
> >  drivers/gpu/drm/i915/intel_ringbuffer.h    | 10 ++---
> >  7 files changed, 64 insertions(+), 61 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
> > index b0fd6a7..8db144b 100644
> > --- a/drivers/gpu/drm/i915/i915_cmd_parser.c
> > +++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
> > @@ -62,23 +62,23 @@
> >   * The parser always rejects such commands.
> >   *
> >   * The majority of the problematic commands fall in the MI_* range, with only a
> > - * few specific commands on each ring (e.g. PIPE_CONTROL and MI_FLUSH_DW).
> > + * few specific commands on each engine (e.g. PIPE_CONTROL and MI_FLUSH_DW).
> >   *
> >   * Implementation:
> > - * Each ring maintains tables of commands and registers which the parser uses in
> > - * scanning batch buffers submitted to that ring.
> > + * Each engine maintains tables of commands and registers which the parser
> > + * uses in scanning batch buffers submitted to that engine.
> >   *
> >   * Since the set of commands that the parser must check for is significantly
> >   * smaller than the number of commands supported, the parser tables contain only
> >   * those commands required by the parser. This generally works because command
> >   * opcode ranges have standard command length encodings. So for commands that
> >   * the parser does not need to check, it can easily skip them. This is
> > - * implemented via a per-ring length decoding vfunc.
> > + * implemented via a per-engine length decoding vfunc.
> >   *
> >   * Unfortunately, there are a number of commands that do not follow the standard
> >   * length encoding for their opcode range, primarily amongst the MI_* commands.
> >   * To handle this, the parser provides a way to define explicit "skip" entries
> > - * in the per-ring command tables.
> > + * in the per-engine command tables.
> >   *
> >   * Other command table entries map fairly directly to high level categories
> >   * mentioned above: rejected, master-only, register whitelist. The parser
> > @@ -603,7 +603,7 @@ static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
> >  	return 0;
> >  }
> >  
> > -static bool validate_cmds_sorted(struct intel_engine_cs *engine,
> > +static bool validate_cmds_sorted(const struct intel_engine_cs *engine,
> 
> This otherwise unrelated change is a dependency from a signature change
> below.
> 
> >  				 const struct drm_i915_cmd_table *cmd_tables,
> >  				 int cmd_table_count)
> >  {
> > @@ -624,8 +624,9 @@ static bool validate_cmds_sorted(struct intel_engine_cs *engine,
> >  			u32 curr = desc->cmd.value & desc->cmd.mask;
> >  
> >  			if (curr < previous) {
> > -				DRM_ERROR("CMD: table not sorted ring=%d table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
> > -					  engine->id, i, j, curr, previous);
> > +				DRM_ERROR("CMD: %s [%d] command table not sorted: table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
> 
> Cut this line here like;
> 
> DRM_ERROR("CMD: %s [%d] command table not sorted: "
> 	  "table=%d entry=%d cmd=0x%08X prev=0x%08X\n"

checkpatch.pl dislikes split strings just as much as it dislikes long
lines, but done.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()

[Joonas Lahtinen]

On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> Since intel_engine_enable_signaling() is now only called via
> fence_enable_sw_signaling(), we can rely on it to provide serialisation
> and run-once for us and so make ourselves slightly simpler.
> 

Originally left this patch for Tvrtko, not remembering he's traveling.

assert_spin_locked() should make this enough robust that it'll still do
what it did perviously.

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

Regards, Joonas

> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/intel_breadcrumbs.c | 13 ++++---------
>  1 file changed, 4 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c
> index d893ccdd62ac..90867446f1a5 100644
> --- a/drivers/gpu/drm/i915/intel_breadcrumbs.c
> +++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c
> @@ -480,19 +480,15 @@ void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
>  	struct rb_node *parent, **p;
>  	bool first, wakeup;
>  
> -	if (unlikely(READ_ONCE(request->signaling.wait.tsk)))
> -		return;
> -
> -	spin_lock(&b->lock);
> -	if (unlikely(request->signaling.wait.tsk)) {
> -		wakeup = false;
> -		goto unlock;
> -	}
> +	/* locked by fence_enable_sw_signaling() */
> +	assert_spin_locked(&request->lock);
>  
>  	request->signaling.wait.tsk = b->signaler;
>  	request->signaling.wait.seqno = request->fence.seqno;
>  	i915_gem_request_get(request);
>  
> +	spin_lock(&b->lock);
> +
>  	/* First add ourselves into the list of waiters, but register our
>  	 * bottom-half as the signaller thread. As per usual, only the oldest
>  	 * waiter (not just signaller) is tasked as the bottom-half waking
> @@ -525,7 +521,6 @@ void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
>  	if (first)
>  		smp_store_mb(b->first_signal, request);
>  
> -unlock:
>  	spin_unlock(&b->lock);
>  
>  	if (wakeup)
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 08/31] drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c

[Dave Gordon]

On 25/07/16 09:49, Chris Wilson wrote:
> On Mon, Jul 25, 2016 at 11:45:42AM +0300, Joonas Lahtinen wrote:
>> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
>>> A few places we use ring when referring to the struct intel_engine_cs. An
>>> anachronism we are pruning out.
>>>
>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>> ---
>>>  drivers/gpu/drm/i915/i915_gem.c | 24 ++++++++++++------------
>>>  1 file changed, 12 insertions(+), 12 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
>>> index e155e8dd28ed..7bfce1d5c61b 100644
>>> --- a/drivers/gpu/drm/i915/i915_gem.c
>>> +++ b/drivers/gpu/drm/i915/i915_gem.c
>>> @@ -46,7 +46,7 @@ static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *o
>>>  static void
>>>  i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
>>>  static void
>>> -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring);
>>> +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int engine);

'engine' rather than 'engine_id' ?
Elsewhere 'engine' is a pointer-to-intel_engine_cs.

>>>
>>>  static bool cpu_cache_is_coherent(struct drm_device *dev,
>>>  				  enum i915_cache_level level)
>>> @@ -1385,10 +1385,10 @@ static void
>>>  i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
>>>  			       struct drm_i915_gem_request *req)
>>>  {
>>> -	int ring = req->engine->id;
>>> +	int idx = req->engine->id;
>>
>> See below.
>>
>>>
>>> -	if (obj->last_read_req[ring] == req)
>>> -		i915_gem_object_retire__read(obj, ring);
>>> +	if (obj->last_read_req[idx] == req)
>>> +		i915_gem_object_retire__read(obj, idx);
>>>  	else if (obj->last_write_req == req)
>>>  		i915_gem_object_retire__write(obj);
>>>
>>> @@ -2381,20 +2381,20 @@ i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
>>>  }
>>>
>>>  static void
>>> -i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
>>> +i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)
>>
>> I do not fancy declaring different variable names than are used. Also,
>> idx is very cryptic in this function signature (one would think of
>> object index).
>
> If you look at the later patches (posted earlier ;) using index is the
> less cryptic option as it really does refer to the index of the tracker.
>
> There was a desired to avoid using ring and here I was trying to avoid
> confusion with activity tracking.
> -Chris

'engine_id' would be unambiguous, although a bit ugly.

I'd rather have the 'id' bit at the end of the name rather than the 
start, but maybe we can find a good abbreviation for the engine_ part of 
the name? It's not a word that contracts easily :(

.Dave.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Dave Gordon]

On 25/07/16 08:44, Chris Wilson wrote:
> Space for flushing the GPU cache prior to completing the request is
> preallocated and so cannot fail.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
>  drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c |  9 +---
>  drivers/gpu/drm/i915/i915_gem_gtt.c        | 11 +++--
>  drivers/gpu/drm/i915/i915_gem_request.c    |  7 ++-
>  drivers/gpu/drm/i915/intel_lrc.c           | 47 +++----------------
>  drivers/gpu/drm/i915/intel_lrc.h           |  2 -
>  drivers/gpu/drm/i915/intel_ringbuffer.c    | 72 +++++++-----------------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h    |  7 ---
>  8 files changed, 37 insertions(+), 120 deletions(-)

[snip]

> -static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
> -{
> -	struct intel_engine_cs *engine = req->engine;
> -	uint32_t flush_domains;
> -	int ret;
> -
> -	flush_domains = 0;
> -	if (engine->gpu_caches_dirty)
> -		flush_domains = I915_GEM_GPU_DOMAINS;
> -
> -	ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
> -	if (ret)
> -		return ret;
> -
> -	engine->gpu_caches_dirty = false;
> -	return 0;
> -}
> -
>  static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>  				 struct list_head *vmas)
>  {
> @@ -690,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>  	/* Unconditionally invalidate gpu caches and ensure that we do flush
>  	 * any residual writes from the previous batch.
>  	 */
> -	return logical_ring_invalidate_all_caches(req);
> +	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
>  }

I don't think the direct call to the vfunc is as clear as to what we're 
trying to achieve here. I'd like some flavour of flush_caches() and 
invalidate_caches() reinstated, even if they're just trivial wrappers 
round the ->emit_flush().

While we're here, could we simplify the parameters? AFAICT we need only 
three permutations: FLUSH (only), INVALIDATE (only) or FLUSH and 
INVALIDATE; and in each case each parameter is either GEM_GPU_DOMAINS or 0.

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Chris Wilson]

On Wed, Jul 27, 2016 at 10:49:46AM +0100, Dave Gordon wrote:
> On 25/07/16 08:44, Chris Wilson wrote:
> >Space for flushing the GPU cache prior to completing the request is
> >preallocated and so cannot fail.
> >
> >Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >---
> > drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
> > drivers/gpu/drm/i915/i915_gem_execbuffer.c |  9 +---
> > drivers/gpu/drm/i915/i915_gem_gtt.c        | 11 +++--
> > drivers/gpu/drm/i915/i915_gem_request.c    |  7 ++-
> > drivers/gpu/drm/i915/intel_lrc.c           | 47 +++----------------
> > drivers/gpu/drm/i915/intel_lrc.h           |  2 -
> > drivers/gpu/drm/i915/intel_ringbuffer.c    | 72 +++++++-----------------------
> > drivers/gpu/drm/i915/intel_ringbuffer.h    |  7 ---
> > 8 files changed, 37 insertions(+), 120 deletions(-)
> 
> [snip]
> 
> >-static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
> >-{
> >-	struct intel_engine_cs *engine = req->engine;
> >-	uint32_t flush_domains;
> >-	int ret;
> >-
> >-	flush_domains = 0;
> >-	if (engine->gpu_caches_dirty)
> >-		flush_domains = I915_GEM_GPU_DOMAINS;
> >-
> >-	ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
> >-	if (ret)
> >-		return ret;
> >-
> >-	engine->gpu_caches_dirty = false;
> >-	return 0;
> >-}
> >-
> > static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
> > 				 struct list_head *vmas)
> > {
> >@@ -690,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
> > 	/* Unconditionally invalidate gpu caches and ensure that we do flush
> > 	 * any residual writes from the previous batch.
> > 	 */
> >-	return logical_ring_invalidate_all_caches(req);
> >+	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
> > }
> 
> I don't think the direct call to the vfunc is as clear as to what
> we're trying to achieve here. I'd like some flavour of
> flush_caches() and invalidate_caches() reinstated, even if they're
> just trivial wrappers round the ->emit_flush().

> While we're here, could we simplify the parameters? AFAICT we need
> only three permutations: FLUSH (only), INVALIDATE (only) or FLUSH
> and INVALIDATE; and in each case each parameter is either
> GEM_GPU_DOMAINS or 0.

Yes, a couple of years ago I sent patches to reduce it down to a single
parameter, (INVALIDATE, FLUSH, BARRIER).

The choice now is which would you prefer

i915_gem_request_emit_flush() {
	req->engine->emit_flush(req, 0, I915_GEM_GPU_DOMAINS);
}
i915_gem_request_emit_invalidate() {
	req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
}

or

	engine->emit_flush(req, INVALIDATE);
	engine->emit_flush(req, FLUSH);

Using the vfunc directly is consistent with elsewhere.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH] drm/i915: Reduce engine->emit_flush() to a single mode parameter

[Chris Wilson]

Rather than passing a complete set of GPU cache domains for either
invalidation or for flushing, or even both, just pass a single parameter
to the engine->emit_flush to determine the required operations.

engine->emit_flush(GPU, 0) -> engine->emit_flush(EMIT_INVALIDATE)
engine->emit_flush(0, GPU) -> engine->emit_flush(EMIT_FLUSH)
engine->emit_flush(GPU, GPU) -> engine->emit_flush(EMIT_FLUSH | EMIT_INVALIDATE)

This allows us to extend the behaviour easily in future, for example if
we want just a command barrier without the overhead of flushing.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Dave Gordon <david.s.gordon@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
---
 drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
 drivers/gpu/drm/i915/i915_gem_gtt.c        | 10 ++----
 drivers/gpu/drm/i915/i915_gem_request.c    |  2 +-
 drivers/gpu/drm/i915/intel_lrc.c           | 23 +++++-------
 drivers/gpu/drm/i915/intel_ringbuffer.c    | 57 +++++++++++-------------------
 drivers/gpu/drm/i915/intel_ringbuffer.h    |  6 ++--
 7 files changed, 38 insertions(+), 64 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index beece8feb8fe..edde8411c478 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -568,7 +568,7 @@ mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags)
 	 * itlb_before_ctx_switch.
 	 */
 	if (IS_GEN6(dev_priv)) {
-		ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
+		ret = engine->emit_flush(req, EMIT_INVALIDATE);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 35c4c595e5ba..e49776e34eed 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -999,7 +999,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 		wmb();
 
 	/* Unconditionally invalidate GPU caches and TLBs. */
-	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
+	return req->engine->emit_flush(req, EMIT_INVALIDATE);
 }
 
 static bool
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 39fa9eb10514..671b1cab5e54 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1666,8 +1666,7 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	int ret;
 
 	/* NB: TLBs must be flushed and invalidated before a switch */
-	ret = engine->emit_flush(req,
-				 I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);
 	if (ret)
 		return ret;
 
@@ -1694,8 +1693,7 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 	int ret;
 
 	/* NB: TLBs must be flushed and invalidated before a switch */
-	ret = engine->emit_flush(req,
-				 I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);
 	if (ret)
 		return ret;
 
@@ -1713,9 +1711,7 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
 
 	/* XXX: RCS is the only one to auto invalidate the TLBs? */
 	if (engine->id != RCS) {
-		ret = engine->emit_flush(req,
-					 I915_GEM_GPU_DOMAINS,
-					 I915_GEM_GPU_DOMAINS);
+		ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 7e3206051ced..67f16feb0552 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -451,7 +451,7 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 * what.
 	 */
 	if (flush_caches) {
-		ret = engine->emit_flush(request, 0, I915_GEM_GPU_DOMAINS);
+		ret = engine->emit_flush(request, EMIT_FLUSH);
 
 		/* Not allowed to fail! */
 		WARN(ret, "engine->emit_flush() failed: %d!\n", ret);
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 409ce5b89d20..b986a3cdba89 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -672,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
 	/* Unconditionally invalidate gpu caches and ensure that we do flush
 	 * any residual writes from the previous batch.
 	 */
-	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
+	return req->engine->emit_flush(req, EMIT_INVALIDATE);
 }
 
 int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
@@ -998,9 +998,7 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (w->count == 0)
 		return 0;
 
-	ret = req->engine->emit_flush(req,
-				      I915_GEM_GPU_DOMAINS,
-				      I915_GEM_GPU_DOMAINS);
+	ret = req->engine->emit_flush(req, EMIT_BARRIER);
 	if (ret)
 		return ret;
 
@@ -1017,9 +1015,7 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
 
 	intel_ring_advance(ring);
 
-	ret = req->engine->emit_flush(req,
-				      I915_GEM_GPU_DOMAINS,
-				      I915_GEM_GPU_DOMAINS);
+	ret = req->engine->emit_flush(req, EMIT_BARRIER);
 	if (ret)
 		return ret;
 
@@ -1600,9 +1596,7 @@ static void gen8_logical_ring_disable_irq(struct intel_engine_cs *engine)
 	I915_WRITE_IMR(engine, ~engine->irq_keep_mask);
 }
 
-static int gen8_emit_flush(struct drm_i915_gem_request *request,
-			   u32 invalidate_domains,
-			   u32 unused)
+static int gen8_emit_flush(struct drm_i915_gem_request *request, u32 mode)
 {
 	struct intel_ring *ring = request->ring;
 	u32 cmd;
@@ -1621,7 +1615,7 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 	 */
 	cmd |= MI_FLUSH_DW_STORE_INDEX | MI_FLUSH_DW_OP_STOREDW;
 
-	if (invalidate_domains & I915_GEM_GPU_DOMAINS) {
+	if (mode & EMIT_INVALIDATE) {
 		cmd |= MI_INVALIDATE_TLB;
 		if (request->engine->id == VCS)
 			cmd |= MI_INVALIDATE_BSD;
@@ -1639,8 +1633,7 @@ static int gen8_emit_flush(struct drm_i915_gem_request *request,
 }
 
 static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
-				  u32 invalidate_domains,
-				  u32 flush_domains)
+				  u32 mode)
 {
 	struct intel_ring *ring = request->ring;
 	struct intel_engine_cs *engine = request->engine;
@@ -1652,14 +1645,14 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 
 	flags |= PIPE_CONTROL_CS_STALL;
 
-	if (flush_domains) {
+	if (mode & EMIT_FLUSH) {
 		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_FLUSH_ENABLE;
 	}
 
-	if (invalidate_domains) {
+	if (mode & EMIT_INVALIDATE) {
 		flags |= PIPE_CONTROL_TLB_INVALIDATE;
 		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 9e4b49644553..1f876e7ce582 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -67,19 +67,15 @@ static void __intel_engine_submit(struct intel_engine_cs *engine)
 }
 
 static int
-gen2_render_ring_flush(struct drm_i915_gem_request *req,
-		       u32	invalidate_domains,
-		       u32	flush_domains)
+gen2_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	u32 cmd;
 	int ret;
 
 	cmd = MI_FLUSH;
-	if (((invalidate_domains|flush_domains) & I915_GEM_DOMAIN_RENDER) == 0)
-		cmd |= MI_NO_WRITE_FLUSH;
 
-	if (invalidate_domains & I915_GEM_DOMAIN_SAMPLER)
+	if (mode & EMIT_INVALIDATE)
 		cmd |= MI_READ_FLUSH;
 
 	ret = intel_ring_begin(req, 2);
@@ -94,9 +90,7 @@ gen2_render_ring_flush(struct drm_i915_gem_request *req,
 }
 
 static int
-gen4_render_ring_flush(struct drm_i915_gem_request *req,
-		       u32	invalidate_domains,
-		       u32	flush_domains)
+gen4_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	u32 cmd;
@@ -131,7 +125,7 @@ gen4_render_ring_flush(struct drm_i915_gem_request *req,
 	 */
 
 	cmd = MI_FLUSH;
-	if (invalidate_domains) {
+	if (mode & EMIT_INVALIDATE) {
 		cmd |= MI_EXE_FLUSH;
 		if (IS_G4X(req->i915) || IS_GEN5(req->i915))
 			cmd |= MI_INVALIDATE_ISP;
@@ -222,8 +216,7 @@ intel_emit_post_sync_nonzero_flush(struct drm_i915_gem_request *req)
 }
 
 static int
-gen6_render_ring_flush(struct drm_i915_gem_request *req,
-		       u32 invalidate_domains, u32 flush_domains)
+gen6_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	u32 scratch_addr =
@@ -240,7 +233,7 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 	 * number of bits based on the write domains has little performance
 	 * impact.
 	 */
-	if (flush_domains) {
+	if (mode & EMIT_FLUSH) {
 		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
 		/*
@@ -249,7 +242,7 @@ gen6_render_ring_flush(struct drm_i915_gem_request *req,
 		 */
 		flags |= PIPE_CONTROL_CS_STALL;
 	}
-	if (invalidate_domains) {
+	if (mode & EMIT_INVALIDATE) {
 		flags |= PIPE_CONTROL_TLB_INVALIDATE;
 		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
@@ -297,8 +290,7 @@ gen7_render_ring_cs_stall_wa(struct drm_i915_gem_request *req)
 }
 
 static int
-gen7_render_ring_flush(struct drm_i915_gem_request *req,
-		       u32 invalidate_domains, u32 flush_domains)
+gen7_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	u32 scratch_addr =
@@ -320,13 +312,13 @@ gen7_render_ring_flush(struct drm_i915_gem_request *req,
 	 * number of bits based on the write domains has little performance
 	 * impact.
 	 */
-	if (flush_domains) {
+	if (mode & EMIT_FLUSH) {
 		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_FLUSH_ENABLE;
 	}
-	if (invalidate_domains) {
+	if (mode & EMIT_INVALIDATE) {
 		flags |= PIPE_CONTROL_TLB_INVALIDATE;
 		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
@@ -384,8 +376,7 @@ gen8_emit_pipe_control(struct drm_i915_gem_request *req,
 }
 
 static int
-gen8_render_ring_flush(struct drm_i915_gem_request *req,
-		       u32 invalidate_domains, u32 flush_domains)
+gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	u32 scratch_addr = req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	u32 flags = 0;
@@ -393,13 +384,13 @@ gen8_render_ring_flush(struct drm_i915_gem_request *req,
 
 	flags |= PIPE_CONTROL_CS_STALL;
 
-	if (flush_domains) {
+	if (mode & EMIT_FLUSH) {
 		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
 		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_FLUSH_ENABLE;
 	}
-	if (invalidate_domains) {
+	if (mode & EMIT_INVALIDATE) {
 		flags |= PIPE_CONTROL_TLB_INVALIDATE;
 		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
@@ -688,9 +679,7 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 	if (w->count == 0)
 		return 0;
 
-	ret = req->engine->emit_flush(req,
-				      I915_GEM_GPU_DOMAINS,
-				      I915_GEM_GPU_DOMAINS);
+	ret = req->engine->emit_flush(req, EMIT_BARRIER);
 	if (ret)
 		return ret;
 
@@ -707,9 +696,7 @@ static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
 
 	intel_ring_advance(ring);
 
-	ret = req->engine->emit_flush(req,
-				      I915_GEM_GPU_DOMAINS,
-				      I915_GEM_GPU_DOMAINS);
+	ret = req->engine->emit_flush(req, EMIT_BARRIER);
 	if (ret)
 		return ret;
 
@@ -1700,9 +1687,7 @@ i8xx_irq_disable(struct intel_engine_cs *engine)
 }
 
 static int
-bsd_ring_flush(struct drm_i915_gem_request *req,
-	       u32     invalidate_domains,
-	       u32     flush_domains)
+bsd_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	int ret;
@@ -2533,8 +2518,7 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
 }
 
-static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
-			       u32 invalidate, u32 flush)
+static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	uint32_t cmd;
@@ -2561,7 +2545,7 @@ static int gen6_bsd_ring_flush(struct drm_i915_gem_request *req,
 	 * operation is complete. This bit is only valid when the
 	 * Post-Sync Operation field is a value of 1h or 3h."
 	 */
-	if (invalidate & I915_GEM_GPU_DOMAINS)
+	if (mode & EMIT_INVALIDATE)
 		cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD;
 
 	intel_ring_emit(ring, cmd);
@@ -2653,8 +2637,7 @@ gen6_ring_dispatch_execbuffer(struct drm_i915_gem_request *req,
 
 /* Blitter support (SandyBridge+) */
 
-static int gen6_ring_flush(struct drm_i915_gem_request *req,
-			   u32 invalidate, u32 flush)
+static int gen6_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
 	struct intel_ring *ring = req->ring;
 	uint32_t cmd;
@@ -2681,7 +2664,7 @@ static int gen6_ring_flush(struct drm_i915_gem_request *req,
 	 * operation is complete. This bit is only valid when the
 	 * Post-Sync Operation field is a value of 1h or 3h."
 	 */
-	if (invalidate & I915_GEM_DOMAIN_RENDER)
+	if (mode & EMIT_INVALIDATE)
 		cmd |= MI_INVALIDATE_TLB;
 	intel_ring_emit(ring, cmd);
 	intel_ring_emit(ring,
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 00723401f98c..76d0495943c3 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -292,8 +292,10 @@ struct intel_engine_cs {
 	u32 ctx_desc_template;
 	int		(*emit_request)(struct drm_i915_gem_request *request);
 	int		(*emit_flush)(struct drm_i915_gem_request *request,
-				      u32 invalidate_domains,
-				      u32 flush_domains);
+				      u32 mode);
+#define EMIT_INVALIDATE	BIT(0)
+#define EMIT_FLUSH	BIT(1)
+#define EMIT_BARRIER	(EMIT_INVALIDATE | EMIT_FLUSH)
 	int		(*emit_bb_start)(struct drm_i915_gem_request *req,
 					 u64 offset, unsigned dispatch_flags);
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Ro.CI.BAT: failure for series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev3)

[Patchwork]

== Series Details ==

Series: series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev3)
URL   : https://patchwork.freedesktop.org/series/10230/
State : failure

== Summary ==

M	drivers/gpu/drm/i915/i915_gem_request.c
M	drivers/gpu/drm/i915/i915_gem_shrinker.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915: Only clear the client pointer when tearing down the file
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915: Only drop the batch-pool's object reference
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem_batch_pool.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915/cmdparser: Remove stray intel_engine_cs *ring
Applying: drm/i915: Use engine to refer to the user's BSD intel_engine_cs
Applying: drm/i915: Avoid using intel_engine_cs *ring for GPU error capture
Applying: drm/i915: Remove stray intel_engine_cs ring identifiers from i915_gem.c
Applying: drm/i915: Update a couple of hangcheck comments to talk about engines
Applying: drm/i915: Unify intel_logical_ring_emit and intel_ring_emit
Applying: drm/i915: Rename request->ringbuf to request->ring
Applying: drm/i915: Rename backpointer from intel_ringbuffer to intel_engine_cs
Applying: drm/i915: Rename intel_context[engine].ringbuf
Applying: drm/i915: Rename struct intel_ringbuffer to struct intel_ring
Applying: drm/i915: Rename residual ringbuf parameters
Applying: drm/i915: Rename intel_pin_and_map_ring()
Applying: drm/i915: Reduce engine->emit_flush() to a single mode parameter
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem_context.c
M	drivers/gpu/drm/i915/i915_gem_execbuffer.c
M	drivers/gpu/drm/i915/i915_gem_gtt.c
M	drivers/gpu/drm/i915/i915_gem_request.c
M	drivers/gpu/drm/i915/intel_lrc.c
M	drivers/gpu/drm/i915/intel_ringbuffer.c
M	drivers/gpu/drm/i915/intel_ringbuffer.h
Falling back to patching base and 3-way merge...
Auto-merging drivers/gpu/drm/i915/intel_ringbuffer.h
Auto-merging drivers/gpu/drm/i915/intel_ringbuffer.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/intel_ringbuffer.c
Auto-merging drivers/gpu/drm/i915/intel_lrc.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/intel_lrc.c
Auto-merging drivers/gpu/drm/i915/i915_gem_request.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/i915_gem_request.c
Auto-merging drivers/gpu/drm/i915/i915_gem_gtt.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/i915_gem_gtt.c
Auto-merging drivers/gpu/drm/i915/i915_gem_execbuffer.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/i915_gem_execbuffer.c
Auto-merging drivers/gpu/drm/i915/i915_gem_context.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/i915_gem_context.c
error: Failed to merge in the changes.
Patch failed at 0017 drm/i915: Reduce engine->emit_flush() to a single mode parameter
The copy of the patch that failed is found in: .git/rebase-apply/patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Dave Gordon]

On 25/07/16 10:18, Joonas Lahtinen wrote:
> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
>> If is simpler and leads to more readable code through the callstack if
>> the allocation returns the allocated struct through the return value.
>>
>> The importance of this is that it no longer looks like we accidentally
>> allocate requests as side-effect of calling certain functions.
>
> Dave seems to have expressed to wish to review this around January, so
> CC'ing him here.
>
> From me,
>
> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>
>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>> ---
>>  drivers/gpu/drm/i915/i915_drv.h            |  3 +-
>>  drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
>>  drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
>>  drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
>>  drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
>>  drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
>>  drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
>>  drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
>>  8 files changed, 79 insertions(+), 140 deletions(-)

The actual code looks fine, but the patch amalgamates several different 
changes, only one  of which is mentioned in the description above.

1. Change the signature of i915_gem_object_sync() and its internal
    subfunction __i915_gem_object_sync() to always require a request
    as input, with corresponding rework of the callsites. This enables
    the removal of the internal call to i915_gem_request_alloc().

2. Change API of i915_gem_request_alloc() not to allow NULL ctx,
    which involves changing various of the remaining callsites.

3. Flatten __i915_gem_request_alloc() into i915_gem_request_alloc().

The description really only covers (3), which is the simplest and most 
mechanical of the changes. So preferably two or three separate patches, 
in whatever order makes the most sense, with a bit more description of each.

.Dave.

>> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
>> index 09e72b0510ab..1031f5a7126a 100644
>> --- a/drivers/gpu/drm/i915/i915_drv.h
>> +++ b/drivers/gpu/drm/i915/i915_drv.h
>> @@ -3171,8 +3171,7 @@ static inline void i915_gem_object_unpin_map(struct drm_i915_gem_object *obj)
>>
>>  int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
>>  int i915_gem_object_sync(struct drm_i915_gem_object *obj,
>> -			 struct intel_engine_cs *to,
>> -			 struct drm_i915_gem_request **to_req);
>> +			 struct drm_i915_gem_request *to);
>>  void i915_vma_move_to_active(struct i915_vma *vma,
>>  			     struct drm_i915_gem_request *req);
>>  int i915_gem_dumb_create(struct drm_file *file_priv,
>> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
>> index 59890f523c5f..b6c4ff63725f 100644
>> --- a/drivers/gpu/drm/i915/i915_gem.c
>> +++ b/drivers/gpu/drm/i915/i915_gem.c
>> @@ -2845,51 +2845,35 @@ out:
>>
>>  static int
>>  __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>> -		       struct intel_engine_cs *to,
>> -		       struct drm_i915_gem_request *from_req,
>> -		       struct drm_i915_gem_request **to_req)
>> +		       struct drm_i915_gem_request *to,
>> +		       struct drm_i915_gem_request *from)
>>  {
>> -	struct intel_engine_cs *from;
>>  	int ret;
>>
>> -	from = i915_gem_request_get_engine(from_req);
>> -	if (to == from)
>> +	if (to->engine == from->engine)
>>  		return 0;
>>
>> -	if (i915_gem_request_completed(from_req))
>> +	if (i915_gem_request_completed(from))
>>  		return 0;
>>
>>  	if (!i915.semaphores) {
>> -		struct drm_i915_private *i915 = to_i915(obj->base.dev);
>> -		ret = __i915_wait_request(from_req,
>> -					  i915->mm.interruptible,
>> +		ret = __i915_wait_request(from,
>> +					  from->i915->mm.interruptible,
>>  					  NULL,
>>  					  NO_WAITBOOST);
>>  		if (ret)
>>  			return ret;
>>
>> -		i915_gem_object_retire_request(obj, from_req);
>> +		i915_gem_object_retire_request(obj, from);
>>  	} else {
>> -		int idx = intel_engine_sync_index(from, to);
>> -		u32 seqno = i915_gem_request_get_seqno(from_req);
>> +		int idx = intel_engine_sync_index(from->engine, to->engine);
>> +		u32 seqno = i915_gem_request_get_seqno(from);
>>
>> -		WARN_ON(!to_req);
>> -
>> -		if (seqno <= from->semaphore.sync_seqno[idx])
>> +		if (seqno <= from->engine->semaphore.sync_seqno[idx])
>>  			return 0;
>>
>> -		if (*to_req == NULL) {
>> -			struct drm_i915_gem_request *req;
>> -
>> -			req = i915_gem_request_alloc(to, NULL);
>> -			if (IS_ERR(req))
>> -				return PTR_ERR(req);
>> -
>> -			*to_req = req;
>> -		}
>> -
>> -		trace_i915_gem_ring_sync_to(*to_req, from, from_req);
>> -		ret = to->semaphore.sync_to(*to_req, from, seqno);
>> +		trace_i915_gem_ring_sync_to(to, from);
>> +		ret = to->engine->semaphore.sync_to(to, from->engine, seqno);
>>  		if (ret)
>>  			return ret;
>>
>> @@ -2897,8 +2881,8 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>>  		 * might have just caused seqno wrap under
>>  		 * the radar.
>>  		 */
>> -		from->semaphore.sync_seqno[idx] =
>> -			i915_gem_request_get_seqno(obj->last_read_req[from->id]);
>> +		from->engine->semaphore.sync_seqno[idx] =
>> +			i915_gem_request_get_seqno(obj->last_read_req[from->engine->id]);
>>  	}
>>
>>  	return 0;
>> @@ -2908,17 +2892,12 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>>   * i915_gem_object_sync - sync an object to a ring.
>>   *
>>   * @obj: object which may be in use on another ring.
>> - * @to: ring we wish to use the object on. May be NULL.
>> - * @to_req: request we wish to use the object for. See below.
>> - *          This will be allocated and returned if a request is
>> - *          required but not passed in.
>> + * @to: request we are wishing to use
>>   *
>>   * This code is meant to abstract object synchronization with the GPU.
>> - * Calling with NULL implies synchronizing the object with the CPU
>> - * rather than a particular GPU ring. Conceptually we serialise writes
>> - * between engines inside the GPU. We only allow one engine to write
>> - * into a buffer at any time, but multiple readers. To ensure each has
>> - * a coherent view of memory, we must:
>> + * Conceptually we serialise writes between engines inside the GPU.
>> + * We only allow one engine to write into a buffer at any time, but
>> + * multiple readers. To ensure each has a coherent view of memory, we must:
>>   *
>>   * - If there is an outstanding write request to the object, the new
>>   *   request must wait for it to complete (either CPU or in hw, requests
>> @@ -2927,22 +2906,11 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
>>   * - If we are a write request (pending_write_domain is set), the new
>>   *   request must wait for outstanding read requests to complete.
>>   *
>> - * For CPU synchronisation (NULL to) no request is required. For syncing with
>> - * rings to_req must be non-NULL. However, a request does not have to be
>> - * pre-allocated. If *to_req is NULL and sync commands will be emitted then a
>> - * request will be allocated automatically and returned through *to_req. Note
>> - * that it is not guaranteed that commands will be emitted (because the system
>> - * might already be idle). Hence there is no need to create a request that
>> - * might never have any work submitted. Note further that if a request is
>> - * returned in *to_req, it is the responsibility of the caller to submit
>> - * that request (after potentially adding more work to it).
>> - *
>>   * Returns 0 if successful, else propagates up the lower layer error.
>>   */
>>  int
>>  i915_gem_object_sync(struct drm_i915_gem_object *obj,
>> -		     struct intel_engine_cs *to,
>> -		     struct drm_i915_gem_request **to_req)
>> +		     struct drm_i915_gem_request *to)
>>  {
>>  	const bool readonly = obj->base.pending_write_domain == 0;
>>  	struct drm_i915_gem_request *req[I915_NUM_ENGINES];
>> @@ -2951,9 +2919,6 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
>>  	if (!obj->active)
>>  		return 0;
>>
>> -	if (to == NULL)
>> -		return i915_gem_object_wait_rendering(obj, readonly);
>> -
>>  	n = 0;
>>  	if (readonly) {
>>  		if (obj->last_write_req)
>> @@ -2964,7 +2929,7 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
>>  				req[n++] = obj->last_read_req[i];
>>  	}
>>  	for (i = 0; i < n; i++) {
>> -		ret = __i915_gem_object_sync(obj, to, req[i], to_req);
>> +		ret = __i915_gem_object_sync(obj, to, req[i]);
>>  		if (ret)
>>  			return ret;
>>  	}
>> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> index 048050176ff9..75d19932847d 100644
>> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> @@ -981,7 +981,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
>>  		struct drm_i915_gem_object *obj = vma->obj;
>>
>>  		if (obj->active & other_rings) {
>> -			ret = i915_gem_object_sync(obj, req->engine, &req);
>> +			ret = i915_gem_object_sync(obj, req);
>>  			if (ret)
>>  				return ret;
>>  		}
>> @@ -1427,7 +1427,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>>  {
>>  	struct drm_i915_private *dev_priv = to_i915(dev);
>>  	struct i915_ggtt *ggtt = &dev_priv->ggtt;
>> -	struct drm_i915_gem_request *req = NULL;
>>  	struct eb_vmas *eb;
>>  	struct drm_i915_gem_object *batch_obj;
>>  	struct drm_i915_gem_exec_object2 shadow_exec_entry;
>> @@ -1615,13 +1614,13 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>>  		params->batch_obj_vm_offset = i915_gem_obj_offset(batch_obj, vm);
>>
>>  	/* Allocate a request for this batch buffer nice and early. */
>> -	req = i915_gem_request_alloc(engine, ctx);
>> -	if (IS_ERR(req)) {
>> -		ret = PTR_ERR(req);
>> +	params->request = i915_gem_request_alloc(engine, ctx);
>> +	if (IS_ERR(params->request)) {
>> +		ret = PTR_ERR(params->request);
>>  		goto err_batch_unpin;
>>  	}
>>
>> -	ret = i915_gem_request_add_to_client(req, file);
>> +	ret = i915_gem_request_add_to_client(params->request, file);
>>  	if (ret)
>>  		goto err_request;
>>
>> @@ -1637,7 +1636,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
>>  	params->dispatch_flags          = dispatch_flags;
>>  	params->batch_obj               = batch_obj;
>>  	params->ctx                     = ctx;
>> -	params->request                 = req;
>>
>>  	ret = dev_priv->gt.execbuf_submit(params, args, &eb->vmas);
>>  err_request:
>> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
>> index 7b772d914e23..995ae99e9cf7 100644
>> --- a/drivers/gpu/drm/i915/i915_gem_request.c
>> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
>> @@ -292,10 +292,21 @@ static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno)
>>  	return 0;
>>  }
>>
>> -static inline int
>> -__i915_gem_request_alloc(struct intel_engine_cs *engine,
>> -			 struct i915_gem_context *ctx,
>> -			 struct drm_i915_gem_request **req_out)
>> +/**
>> + * i915_gem_request_alloc - allocate a request structure
>> + *
>> + * @engine: engine that we wish to issue the request on.
>> + * @ctx: context that the request will be associated with.
>> + *       This can be NULL if the request is not directly related to
>> + *       any specific user context, in which case this function will
>> + *       choose an appropriate context to use.
>> + *
>> + * Returns a pointer to the allocated request if successful,
>> + * or an error code if not.
>> + */
>> +struct drm_i915_gem_request *
>> +i915_gem_request_alloc(struct intel_engine_cs *engine,
>> +		       struct i915_gem_context *ctx)
>>  {
>>  	struct drm_i915_private *dev_priv = engine->i915;
>>  	unsigned int reset_counter = i915_reset_counter(&dev_priv->gpu_error);
>> @@ -303,18 +314,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>>  	u32 seqno;
>>  	int ret;
>>
>> -	if (!req_out)
>> -		return -EINVAL;
>> -
>> -	*req_out = NULL;
>> -
>>  	/* ABI: Before userspace accesses the GPU (e.g. execbuffer), report
>>  	 * EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex
>>  	 * and restart.
>>  	 */
>>  	ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible);
>>  	if (ret)
>> -		return ret;
>> +		return ERR_PTR(ret);
>>
>>  	/* Move the oldest request to the slab-cache (if not in use!) */
>>  	req = list_first_entry_or_null(&engine->request_list,
>> @@ -324,7 +330,7 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>>
>>  	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
>>  	if (!req)
>> -		return -ENOMEM;
>> +		return ERR_PTR(-ENOMEM);
>>
>>  	ret = i915_gem_get_seqno(dev_priv, &seqno);
>>  	if (ret)
>> @@ -357,39 +363,13 @@ __i915_gem_request_alloc(struct intel_engine_cs *engine,
>>  	if (ret)
>>  		goto err_ctx;
>>
>> -	*req_out = req;
>> -	return 0;
>> +	return req;
>>
>>  err_ctx:
>>  	i915_gem_context_put(ctx);
>>  err:
>>  	kmem_cache_free(dev_priv->requests, req);
>> -	return ret;
>> -}
>> -
>> -/**
>> - * i915_gem_request_alloc - allocate a request structure
>> - *
>> - * @engine: engine that we wish to issue the request on.
>> - * @ctx: context that the request will be associated with.
>> - *       This can be NULL if the request is not directly related to
>> - *       any specific user context, in which case this function will
>> - *       choose an appropriate context to use.
>> - *
>> - * Returns a pointer to the allocated request if successful,
>> - * or an error code if not.
>> - */
>> -struct drm_i915_gem_request *
>> -i915_gem_request_alloc(struct intel_engine_cs *engine,
>> -		       struct i915_gem_context *ctx)
>> -{
>> -	struct drm_i915_gem_request *req;
>> -	int err;
>> -
>> -	if (!ctx)
>> -		ctx = engine->i915->kernel_context;
>> -	err = __i915_gem_request_alloc(engine, ctx, &req);
>> -	return err ? ERR_PTR(err) : req;
>> +	return ERR_PTR(ret);
>>  }
>>
>>  static void i915_gem_mark_busy(const struct intel_engine_cs *engine)
>> diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
>> index 007112d1e049..9e43c0aa6e3b 100644
>> --- a/drivers/gpu/drm/i915/i915_trace.h
>> +++ b/drivers/gpu/drm/i915/i915_trace.h
>> @@ -449,10 +449,9 @@ TRACE_EVENT(i915_gem_evict_vm,
>>  );
>>
>>  TRACE_EVENT(i915_gem_ring_sync_to,
>> -	    TP_PROTO(struct drm_i915_gem_request *to_req,
>> -		     struct intel_engine_cs *from,
>> -		     struct drm_i915_gem_request *req),
>> -	    TP_ARGS(to_req, from, req),
>> +	    TP_PROTO(struct drm_i915_gem_request *to,
>> +		     struct drm_i915_gem_request *from),
>> +	    TP_ARGS(to, from),
>>
>>  	    TP_STRUCT__entry(
>>  			     __field(u32, dev)
>> @@ -463,9 +462,9 @@ TRACE_EVENT(i915_gem_ring_sync_to,
>>
>>  	    TP_fast_assign(
>>  			   __entry->dev = from->i915->drm.primary->index;
>> -			   __entry->sync_from = from->id;
>> -			   __entry->sync_to = to_req->engine->id;
>> -			   __entry->seqno = req->fence.seqno;
>> +			   __entry->sync_from = from->engine->id;
>> +			   __entry->sync_to = to->engine->id;
>> +			   __entry->seqno = from->fence.seqno;
>>  			   ),
>>
>>  	    TP_printk("dev=%u, sync-from=%u, sync-to=%u, seqno=%u",
>> diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
>> index bff172c45ff7..5d4420b67642 100644
>> --- a/drivers/gpu/drm/i915/intel_display.c
>> +++ b/drivers/gpu/drm/i915/intel_display.c
>> @@ -11583,7 +11583,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>>  	struct intel_flip_work *work;
>>  	struct intel_engine_cs *engine;
>>  	bool mmio_flip;
>> -	struct drm_i915_gem_request *request = NULL;
>> +	struct drm_i915_gem_request *request;
>>  	int ret;
>>
>>  	/*
>> @@ -11690,22 +11690,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>>
>>  	mmio_flip = use_mmio_flip(engine, obj);
>>
>> -	/* When using CS flips, we want to emit semaphores between rings.
>> -	 * However, when using mmio flips we will create a task to do the
>> -	 * synchronisation, so all we want here is to pin the framebuffer
>> -	 * into the display plane and skip any waits.
>> -	 */
>> -	if (!mmio_flip) {
>> -		ret = i915_gem_object_sync(obj, engine, &request);
>> -		if (!ret && !request) {
>> -			request = i915_gem_request_alloc(engine, NULL);
>> -			ret = PTR_ERR_OR_ZERO(request);
>> -		}
>> -
>> -		if (ret)
>> -			goto cleanup_pending;
>> -	}
>> -
>>  	ret = intel_pin_and_fence_fb_obj(fb, primary->state->rotation);
>>  	if (ret)
>>  		goto cleanup_pending;
>> @@ -11723,14 +11707,24 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>>
>>  		schedule_work(&work->mmio_work);
>>  	} else {
>> -		i915_gem_request_assign(&work->flip_queued_req, request);
>> +		request = i915_gem_request_alloc(engine, engine->last_context);
>> +		if (IS_ERR(request)) {
>> +			ret = PTR_ERR(request);
>> +			goto cleanup_unpin;
>> +		}
>> +
>> +		ret = i915_gem_object_sync(obj, request);
>> +		if (ret)
>> +			goto cleanup_request;
>> +
>>  		ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, request,
>>  						   page_flip_flags);
>>  		if (ret)
>> -			goto cleanup_unpin;
>> +			goto cleanup_request;
>>
>>  		intel_mark_page_flip_active(intel_crtc, work);
>>
>> +		work->flip_queued_req = i915_gem_request_get(request);
>>  		i915_add_request_no_flush(request);
>>  	}
>>
>> @@ -11745,11 +11739,11 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
>>
>>  	return 0;
>>
>> +cleanup_request:
>> +	i915_add_request_no_flush(request);
>>  cleanup_unpin:
>>  	intel_unpin_fb_obj(fb, crtc->primary->state->rotation);
>>  cleanup_pending:
>> -	if (!IS_ERR_OR_NULL(request))
>> -		i915_add_request_no_flush(request);
>>  	atomic_dec(&intel_crtc->unpin_work_count);
>>  	mutex_unlock(&dev->struct_mutex);
>>  cleanup:
>> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
>> index 6487794ac96c..a37af1533618 100644
>> --- a/drivers/gpu/drm/i915/intel_lrc.c
>> +++ b/drivers/gpu/drm/i915/intel_lrc.c
>> @@ -655,7 +655,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>>  		struct drm_i915_gem_object *obj = vma->obj;
>>
>>  		if (obj->active & other_rings) {
>> -			ret = i915_gem_object_sync(obj, req->engine, &req);
>> +			ret = i915_gem_object_sync(obj, req);
>>  			if (ret)
>>  				return ret;
>>  		}
>> diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
>> index 8f1d4d9ef345..651efe4e468e 100644
>> --- a/drivers/gpu/drm/i915/intel_overlay.c
>> +++ b/drivers/gpu/drm/i915/intel_overlay.c
>> @@ -229,11 +229,18 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
>>  	return 0;
>>  }
>>
>> +static struct drm_i915_gem_request *alloc_request(struct intel_overlay *overlay)
>> +{
>> +	struct drm_i915_private *dev_priv = overlay->i915;
>> +	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>> +
>> +	return i915_gem_request_alloc(engine, dev_priv->kernel_context);
>> +}
>> +
>>  /* overlay needs to be disable in OCMD reg */
>>  static int intel_overlay_on(struct intel_overlay *overlay)
>>  {
>>  	struct drm_i915_private *dev_priv = overlay->i915;
>> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>>  	struct drm_i915_gem_request *req;
>>  	struct intel_ring *ring;
>>  	int ret;
>> @@ -241,7 +248,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
>>  	WARN_ON(overlay->active);
>>  	WARN_ON(IS_I830(dev_priv) && !(dev_priv->quirks & QUIRK_PIPEA_FORCE));
>>
>> -	req = i915_gem_request_alloc(engine, NULL);
>> +	req = alloc_request(overlay);
>>  	if (IS_ERR(req))
>>  		return PTR_ERR(req);
>>
>> @@ -268,7 +275,6 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
>>  				  bool load_polyphase_filter)
>>  {
>>  	struct drm_i915_private *dev_priv = overlay->i915;
>> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>>  	struct drm_i915_gem_request *req;
>>  	struct intel_ring *ring;
>>  	u32 flip_addr = overlay->flip_addr;
>> @@ -285,7 +291,7 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
>>  	if (tmp & (1 << 17))
>>  		DRM_DEBUG("overlay underrun, DOVSTA: %x\n", tmp);
>>
>> -	req = i915_gem_request_alloc(engine, NULL);
>> +	req = alloc_request(overlay);
>>  	if (IS_ERR(req))
>>  		return PTR_ERR(req);
>>
>> @@ -338,7 +344,6 @@ static void intel_overlay_off_tail(struct intel_overlay *overlay)
>>  static int intel_overlay_off(struct intel_overlay *overlay)
>>  {
>>  	struct drm_i915_private *dev_priv = overlay->i915;
>> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>>  	struct drm_i915_gem_request *req;
>>  	struct intel_ring *ring;
>>  	u32 flip_addr = overlay->flip_addr;
>> @@ -352,7 +357,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
>>  	 * of the hw. Do it in both cases */
>>  	flip_addr |= OFC_UPDATE;
>>
>> -	req = i915_gem_request_alloc(engine, NULL);
>> +	req = alloc_request(overlay);
>>  	if (IS_ERR(req))
>>  		return PTR_ERR(req);
>>
>> @@ -412,7 +417,6 @@ static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay)
>>  static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
>>  {
>>  	struct drm_i915_private *dev_priv = overlay->i915;
>> -	struct intel_engine_cs *engine = &dev_priv->engine[RCS];
>>  	int ret;
>>
>>  	lockdep_assert_held(&dev_priv->drm.struct_mutex);
>> @@ -428,7 +432,7 @@ static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
>>  		struct drm_i915_gem_request *req;
>>  		struct intel_ring *ring;
>>
>> -		req = i915_gem_request_alloc(engine, NULL);
>> +		req = alloc_request(overlay);
>>  		if (IS_ERR(req))
>>  			return PTR_ERR(req);
>>

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Dave Gordon]

On 27/07/16 11:00, Chris Wilson wrote:
> On Wed, Jul 27, 2016 at 10:49:46AM +0100, Dave Gordon wrote:
>> On 25/07/16 08:44, Chris Wilson wrote:
>>> Space for flushing the GPU cache prior to completing the request is
>>> preallocated and so cannot fail.
>>>
>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>> ---
>>> drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
>>> drivers/gpu/drm/i915/i915_gem_execbuffer.c |  9 +---
>>> drivers/gpu/drm/i915/i915_gem_gtt.c        | 11 +++--
>>> drivers/gpu/drm/i915/i915_gem_request.c    |  7 ++-
>>> drivers/gpu/drm/i915/intel_lrc.c           | 47 +++----------------
>>> drivers/gpu/drm/i915/intel_lrc.h           |  2 -
>>> drivers/gpu/drm/i915/intel_ringbuffer.c    | 72 +++++++-----------------------
>>> drivers/gpu/drm/i915/intel_ringbuffer.h    |  7 ---
>>> 8 files changed, 37 insertions(+), 120 deletions(-)
>>
>> [snip]
>>
>>> -static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
>>> -{
>>> -	struct intel_engine_cs *engine = req->engine;
>>> -	uint32_t flush_domains;
>>> -	int ret;
>>> -
>>> -	flush_domains = 0;
>>> -	if (engine->gpu_caches_dirty)
>>> -		flush_domains = I915_GEM_GPU_DOMAINS;
>>> -
>>> -	ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
>>> -	if (ret)
>>> -		return ret;
>>> -
>>> -	engine->gpu_caches_dirty = false;
>>> -	return 0;
>>> -}
>>> -
>>> static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>>> 				 struct list_head *vmas)
>>> {
>>> @@ -690,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
>>> 	/* Unconditionally invalidate gpu caches and ensure that we do flush
>>> 	 * any residual writes from the previous batch.
>>> 	 */
>>> -	return logical_ring_invalidate_all_caches(req);
>>> +	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
>>> }
>>
>> I don't think the direct call to the vfunc is as clear as to what
>> we're trying to achieve here. I'd like some flavour of
>> flush_caches() and invalidate_caches() reinstated, even if they're
>> just trivial wrappers round the ->emit_flush().
>
>> While we're here, could we simplify the parameters? AFAICT we need
>> only three permutations: FLUSH (only), INVALIDATE (only) or FLUSH
>> and INVALIDATE; and in each case each parameter is either
>> GEM_GPU_DOMAINS or 0.
>
> Yes, a couple of years ago I sent patches to reduce it down to a single
> parameter, (INVALIDATE, FLUSH, BARRIER).
>
> The choice now is which would you prefer
>
> i915_gem_request_emit_flush() {
> 	req->engine->emit_flush(req, 0, I915_GEM_GPU_DOMAINS);
> }
> i915_gem_request_emit_invalidate() {
> 	req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
> }
>
> or
>
> 	engine->emit_flush(req, INVALIDATE);
> 	engine->emit_flush(req, FLUSH);
>
> Using the vfunc directly is consistent with elsewhere.
> -Chris

I don't mind the naked vfunc if the call looks simple enough, as the 
latter does. It's when the vfunc parameters aren't sufficiently simple 
and obvious that it needs wrapping up.

So let's go with the second option :)

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 17/31] drm/i915: Remove obsolete engine->gpu_caches_dirty

[Joonas Lahtinen]

On ke, 2016-07-27 at 11:00 +0100, Chris Wilson wrote:
> On Wed, Jul 27, 2016 at 10:49:46AM +0100, Dave Gordon wrote:
> > 
> > On 25/07/16 08:44, Chris Wilson wrote:
> > > 
> > > Space for flushing the GPU cache prior to completing the request is
> > > preallocated and so cannot fail.
> > > 
> > > Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> > > ---
> > > drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
> > > drivers/gpu/drm/i915/i915_gem_execbuffer.c |  9 +---
> > > drivers/gpu/drm/i915/i915_gem_gtt.c        | 11 +++--
> > > drivers/gpu/drm/i915/i915_gem_request.c    |  7 ++-
> > > drivers/gpu/drm/i915/intel_lrc.c           | 47 +++----------------
> > > drivers/gpu/drm/i915/intel_lrc.h           |  2 -
> > > drivers/gpu/drm/i915/intel_ringbuffer.c    | 72 +++++++-----------------------
> > > drivers/gpu/drm/i915/intel_ringbuffer.h    |  7 ---
> > > 8 files changed, 37 insertions(+), 120 deletions(-)
> > [snip]
> > 
> > > 
> > > -static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
> > > -{
> > > -	struct intel_engine_cs *engine = req->engine;
> > > -	uint32_t flush_domains;
> > > -	int ret;
> > > -
> > > -	flush_domains = 0;
> > > -	if (engine->gpu_caches_dirty)
> > > -		flush_domains = I915_GEM_GPU_DOMAINS;
> > > -
> > > -	ret = engine->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
> > > -	if (ret)
> > > -		return ret;
> > > -
> > > -	engine->gpu_caches_dirty = false;
> > > -	return 0;
> > > -}
> > > -
> > > static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
> > > 				 struct list_head *vmas)
> > > {
> > > @@ -690,7 +672,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
> > > 	/* Unconditionally invalidate gpu caches and ensure that we do flush
> > > 	 * any residual writes from the previous batch.
> > > 	 */
> > > -	return logical_ring_invalidate_all_caches(req);
> > > +	return req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
> > > }
> > I don't think the direct call to the vfunc is as clear as to what
> > we're trying to achieve here. I'd like some flavour of
> > flush_caches() and invalidate_caches() reinstated, even if they're
> > just trivial wrappers round the ->emit_flush().
> > 
> > While we're here, could we simplify the parameters? AFAICT we need
> > only three permutations: FLUSH (only), INVALIDATE (only) or FLUSH
> > and INVALIDATE; and in each case each parameter is either
> > GEM_GPU_DOMAINS or 0.
> Yes, a couple of years ago I sent patches to reduce it down to a single
> parameter, (INVALIDATE, FLUSH, BARRIER).
> 
> The choice now is which would you prefer
> 
> i915_gem_request_emit_flush() {
> 	req->engine->emit_flush(req, 0, I915_GEM_GPU_DOMAINS);
> }
> i915_gem_request_emit_invalidate() {
> 	req->engine->emit_flush(req, I915_GEM_GPU_DOMAINS, 0);
> }
> 
> or
> 
> 	engine->emit_flush(req, INVALIDATE);
> 	engine->emit_flush(req, FLUSH);

+1 on latter, much more informative.

Regards, Joonas

> 
> Using the vfunc directly is consistent with elsewhere.
> -Chris
> 
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Dave Gordon]

On 25/07/16 08:44, Chris Wilson wrote:
> If we rewrite the I915_WRITE_TAIL specialisation for the legacy
> ringbuffer as submitting the request onto the ringbuffer, we can unify
> the vfunc with both execlists and GuC in the next patch.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/i915_gem_request.c |  8 ++---
>  drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 53 +++++++++++++++++----------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h |  3 +-
>  4 files changed, 32 insertions(+), 34 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> index 1c185e293bf0..8814e9c5266b 100644
> --- a/drivers/gpu/drm/i915/i915_gem_request.c
> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
> @@ -467,15 +467,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>  	 */
>  	request->postfix = ring->tail;
>
> -	if (i915.enable_execlists) {
> +	if (i915.enable_execlists)
>  		ret = engine->emit_request(request);
> -	} else {
> +	else
>  		ret = engine->add_request(request);
> -
> -		request->tail = ring->tail;
> -	}
>  	/* Not allowed to fail! */
>  	WARN(ret, "emit|add_request failed: %d!\n", ret);
> +
>  	/* Sanity check that the reserved size was large enough. */
>  	ret = ring->tail - request_start;
>  	if (ret < 0)
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index 567d94de3300..250edb2bcef7 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
>  	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
>  	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
>
> -	reg_state[CTX_RING_TAIL+1] = rq->tail;
> +	reg_state[CTX_RING_TAIL+1] = rq->tail % (rq->ring->size - 1);

mod ringsize-1 ?

Surely tail % ringsize, or tail & (ringsize-1).

But it's redundant anyway, rq->tail cannot exceed ring->size,
so the original code was correct.

>  	/* True 32b PPGTT with dynamic page allocation: update PDP
>  	 * registers and point the unallocated PDPs to scratch page.
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 0a2e25e67450..3e1049c972e0 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -58,14 +58,6 @@ void intel_ring_update_space(struct intel_ring *ring)
>  					 ring->tail, ring->size);
>  }
>
> -static void __intel_engine_submit(struct intel_engine_cs *engine)
> -{
> -	struct intel_ring *ring = engine->buffer;
> -
> -	ring->tail &= ring->size - 1;
> -	engine->write_tail(engine, ring->tail);
> -}
> -
>  static int
>  gen2_render_ring_flush(struct drm_i915_gem_request *req,
>  		       u32	invalidate_domains,
> @@ -421,13 +413,6 @@ gen8_render_ring_flush(struct drm_i915_gem_request *req,
>  	return gen8_emit_pipe_control(req, flags, scratch_addr);
>  }
>
> -static void ring_write_tail(struct intel_engine_cs *engine,
> -			    u32 value)
> -{
> -	struct drm_i915_private *dev_priv = engine->i915;
> -	I915_WRITE_TAIL(engine, value);
> -}
> -
>  u64 intel_engine_get_active_head(struct intel_engine_cs *engine)
>  {
>  	struct drm_i915_private *dev_priv = engine->i915;
> @@ -541,7 +526,7 @@ static bool stop_ring(struct intel_engine_cs *engine)
>
>  	I915_WRITE_CTL(engine, 0);
>  	I915_WRITE_HEAD(engine, 0);
> -	engine->write_tail(engine, 0);
> +	I915_WRITE_TAIL(engine, 0);
>
>  	if (!IS_GEN2(dev_priv)) {
>  		(void)I915_READ_CTL(engine);
> @@ -1482,7 +1467,10 @@ gen6_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
>  	intel_ring_emit(ring, req->fence.seqno);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
> -	__intel_engine_submit(engine);
> +	intel_ring_advance(ring);
> +
> +	req->tail = ring->tail;
> +	engine->submit_request(req);
>
>  	return 0;
>  }
> @@ -1512,7 +1500,9 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_emit(ring, 0);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
>  	intel_ring_emit(ring, MI_NOOP);
> -	__intel_engine_submit(engine);
> +
> +	req->tail = ring->tail;
> +	engine->submit_request(req);
>
>  	return 0;
>  }
> @@ -1731,11 +1721,22 @@ i9xx_add_request(struct drm_i915_gem_request *req)
>  	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
>  	intel_ring_emit(ring, req->fence.seqno);
>  	intel_ring_emit(ring, MI_USER_INTERRUPT);
> -	__intel_engine_submit(req->engine);
> +	intel_ring_advance(ring);
> +
> +	req->tail = ring->tail;
> +	req->engine->submit_request(req);
>
>  	return 0;
>  }
>
> +static void i9xx_submit_request(struct drm_i915_gem_request *request)
> +{
> +	struct drm_i915_private *dev_priv = request->i915;
> +
> +	I915_WRITE_TAIL(request->engine,
> +			request->tail % (request->ring->size - 1));
> +}

Another "modulo ringsize-1" here too.

.Dave.

>  static void
>  gen6_irq_enable(struct intel_engine_cs *engine)
>  {
> @@ -2494,10 +2495,9 @@ void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
>  	rcu_read_unlock();
>  }
>
> -static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
> -				     u32 value)
> +static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
>  {
> -	struct drm_i915_private *dev_priv = engine->i915;
> +	struct drm_i915_private *dev_priv = request->i915;
>
>  	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
>
> @@ -2521,8 +2521,9 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
>  		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
>
>  	/* Now that the ring is fully powered up, update the tail */
> -	I915_WRITE_FW(RING_TAIL(engine->mmio_base), value);
> -	POSTING_READ_FW(RING_TAIL(engine->mmio_base));
> +	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
> +		      request->tail % (request->ring->size - 1));
> +	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
>
>  	/* Let the ring send IDLE messages to the GT again,
>  	 * and so let it sleep to conserve power when idle.
> @@ -2828,7 +2829,7 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
>  				      struct intel_engine_cs *engine)
>  {
>  	engine->init_hw = init_ring_common;
> -	engine->write_tail = ring_write_tail;
> +	engine->submit_request = i9xx_submit_request;
>
>  	engine->add_request = i9xx_add_request;
>  	if (INTEL_GEN(dev_priv) >= 6)
> @@ -2912,7 +2913,7 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
>  	if (INTEL_GEN(dev_priv) >= 6) {
>  		/* gen6 bsd needs a special wa for tail updates */
>  		if (IS_GEN6(dev_priv))
> -			engine->write_tail = gen6_bsd_ring_write_tail;
> +			engine->submit_request = gen6_bsd_submit_request;
>  		engine->emit_flush = gen6_bsd_ring_flush;
>  		if (INTEL_GEN(dev_priv) < 8)
>  			engine->irq_enable_mask = GT_BSD_USER_INTERRUPT;
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 13b816f2d264..5428a3c288d5 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -204,8 +204,6 @@ struct intel_engine_cs {
>
>  	int		(*init_context)(struct drm_i915_gem_request *req);
>
> -	void		(*write_tail)(struct intel_engine_cs *engine,
> -				      u32 value);
>  	int		(*add_request)(struct drm_i915_gem_request *req);
>  	/* Some chipsets are not quite as coherent as advertised and need
>  	 * an expensive kick to force a true read of the up-to-date seqno.
> @@ -294,6 +292,7 @@ struct intel_engine_cs {
>  #define I915_DISPATCH_SECURE 0x1
>  #define I915_DISPATCH_PINNED 0x2
>  #define I915_DISPATCH_RS     0x4
> +	void		(*submit_request)(struct drm_i915_gem_request *req);
>
>  	/**
>  	 * List of objects currently involved in rendering from the
>

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> On 25/07/16 08:44, Chris Wilson wrote:
> >If we rewrite the I915_WRITE_TAIL specialisation for the legacy
> >ringbuffer as submitting the request onto the ringbuffer, we can unify
> >the vfunc with both execlists and GuC in the next patch.
> >
> >Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> >---
> > drivers/gpu/drm/i915/i915_gem_request.c |  8 ++---
> > drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
> > drivers/gpu/drm/i915/intel_ringbuffer.c | 53 +++++++++++++++++----------------
> > drivers/gpu/drm/i915/intel_ringbuffer.h |  3 +-
> > 4 files changed, 32 insertions(+), 34 deletions(-)
> >
> >diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> >index 1c185e293bf0..8814e9c5266b 100644
> >--- a/drivers/gpu/drm/i915/i915_gem_request.c
> >+++ b/drivers/gpu/drm/i915/i915_gem_request.c
> >@@ -467,15 +467,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
> > 	 */
> > 	request->postfix = ring->tail;
> >
> >-	if (i915.enable_execlists) {
> >+	if (i915.enable_execlists)
> > 		ret = engine->emit_request(request);
> >-	} else {
> >+	else
> > 		ret = engine->add_request(request);
> >-
> >-		request->tail = ring->tail;
> >-	}
> > 	/* Not allowed to fail! */
> > 	WARN(ret, "emit|add_request failed: %d!\n", ret);
> >+
> > 	/* Sanity check that the reserved size was large enough. */
> > 	ret = ring->tail - request_start;
> > 	if (ret < 0)
> >diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> >index 567d94de3300..250edb2bcef7 100644
> >--- a/drivers/gpu/drm/i915/intel_lrc.c
> >+++ b/drivers/gpu/drm/i915/intel_lrc.c
> >@@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
> > 	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
> > 	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
> >
> >-	reg_state[CTX_RING_TAIL+1] = rq->tail;
> >+	reg_state[CTX_RING_TAIL+1] = rq->tail % (rq->ring->size - 1);
> 
> mod ringsize-1 ?
> 
> Surely tail % ringsize, or tail & (ringsize-1).
> 
> But it's redundant anyway, rq->tail cannot exceed ring->size,
> so the original code was correct.

No, rq->tail can be equal to ring->size which leads to a GPU hang.
(Observed on the older gen at least, I'd rather have the same paranoia
here.)
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> On 25/07/16 08:44, Chris Wilson wrote:
> >+static void i9xx_submit_request(struct drm_i915_gem_request *request)
> >+{
> >+	struct drm_i915_private *dev_priv = request->i915;
> >+
> >+	I915_WRITE_TAIL(request->engine,
> >+			request->tail % (request->ring->size - 1));
> >+}
> 
> Another "modulo ringsize-1" here too.

Because the original code included that safe guard against known GPU
hangs.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Chris Wilson]

On Wed, Jul 27, 2016 at 12:08:35PM +0100, Dave Gordon wrote:
> On 25/07/16 10:18, Joonas Lahtinen wrote:
> >On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> >>If is simpler and leads to more readable code through the callstack if
> >>the allocation returns the allocated struct through the return value.
> >>
> >>The importance of this is that it no longer looks like we accidentally
> >>allocate requests as side-effect of calling certain functions.
> >
> >Dave seems to have expressed to wish to review this around January, so
> >CC'ing him here.
> >
> >From me,
> >
> >Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> >
> >>Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >>---
> >> drivers/gpu/drm/i915/i915_drv.h            |  3 +-
> >> drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
> >> drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
> >> drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
> >> drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
> >> drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
> >> drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
> >> drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
> >> 8 files changed, 79 insertions(+), 140 deletions(-)
> 
> The actual code looks fine, but the patch amalgamates several
> different changes, only one  of which is mentioned in the
> description above.
> 
> 1. Change the signature of i915_gem_object_sync() and its internal
>    subfunction __i915_gem_object_sync() to always require a request
>    as input, with corresponding rework of the callsites. This enables
>    the removal of the internal call to i915_gem_request_alloc().
> 
> 2. Change API of i915_gem_request_alloc() not to allow NULL ctx,
>    which involves changing various of the remaining callsites.

Reverts the unwanted the change.

> 3. Flatten __i915_gem_request_alloc() into i915_gem_request_alloc().
> 
> The description really only covers (3), which is the simplest and
> most mechanical of the changes. So preferably two or three separate
> patches, in whatever order makes the most sense, with a bit more
> description of each.

It mentions 1 explicitly which was the reason why the API was originally
intended to be this.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Joonas Lahtinen]

On ke, 2016-07-27 at 13:30 +0100, Chris Wilson wrote:
> On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> > 
> > On 25/07/16 08:44, Chris Wilson wrote:
> > > 
> > > +static void i9xx_submit_request(struct drm_i915_gem_request *request)
> > > +{
> > > +	struct drm_i915_private *dev_priv = request->i915;
> > > +
> > > +	I915_WRITE_TAIL(request->engine,
> > > +			request->tail % (request->ring->size - 1));
> > > +}
> > Another "modulo ringsize-1" here too.
> Because the original code included that safe guard against known GPU
> hangs.

I think a comment would be justified.

Regards, Joonas

> -Chris
> 
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Reduce engine->emit_flush() to a single mode parameter

[Joonas Lahtinen]

On ke, 2016-07-27 at 11:53 +0100, Chris Wilson wrote:
> @@ -1713,9 +1711,7 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
>  
>  	/* XXX: RCS is the only one to auto invalidate the TLBs? */
>  	if (engine->id != RCS) {
> -		ret = engine->emit_flush(req,
> -					 I915_GEM_GPU_DOMAINS,
> -					 I915_GEM_GPU_DOMAINS);
> +		ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);

Not quite sure why you want to use a combination of flags vs. _BARRIER
at different places?

But the code is now much more readable,

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

On Thu, Jul 28, 2016 at 09:41:14AM +0300, Joonas Lahtinen wrote:
> On ke, 2016-07-27 at 13:30 +0100, Chris Wilson wrote:
> > On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> > > 
> > > On 25/07/16 08:44, Chris Wilson wrote:
> > > > 
> > > > +static void i9xx_submit_request(struct drm_i915_gem_request *request)
> > > > +{
> > > > +	struct drm_i915_private *dev_priv = request->i915;
> > > > +
> > > > +	I915_WRITE_TAIL(request->engine,
> > > > +			request->tail % (request->ring->size - 1));
> > > > +}
> > > Another "modulo ringsize-1" here too.
> > Because the original code included that safe guard against known GPU
> > hangs.
> 
> I think a comment would be justified.

Hmm, I do beg your pardon, at this point in the tree
intel_ring_advance() still exists and is applying the modulus on every
command packet.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Joonas Lahtinen]

On to, 2016-07-28 at 08:12 +0100, Chris Wilson wrote:
> On Thu, Jul 28, 2016 at 09:41:14AM +0300, Joonas Lahtinen wrote:
> > 
> > On ke, 2016-07-27 at 13:30 +0100, Chris Wilson wrote:
> > > 
> > > On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> > > > 
> > > > 
> > > > On 25/07/16 08:44, Chris Wilson wrote:
> > > > > 
> > > > > 
> > > > > +static void i9xx_submit_request(struct drm_i915_gem_request *request)
> > > > > +{
> > > > > +	struct drm_i915_private *dev_priv = request->i915;
> > > > > +
> > > > > +	I915_WRITE_TAIL(request->engine,
> > > > > +			request->tail % (request->ring->size - 1));
> > > > > +}
> > > > Another "modulo ringsize-1" here too.
> > > Because the original code included that safe guard against known GPU
> > > hangs.
> > I think a comment would be justified.
> Hmm, I do beg your pardon, at this point in the tree
> intel_ring_advance() still exists and is applying the modulus on every
> command packet.

I'd add a comment there too. Just to reduce confusion in future. Or
provide an unified modulo wrapper with the comment.

Regards, Joonas

> -Chris
> 
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Reduce engine->emit_flush() to a single mode parameter

[Chris Wilson]

On Thu, Jul 28, 2016 at 10:11:43AM +0300, Joonas Lahtinen wrote:
> On ke, 2016-07-27 at 11:53 +0100, Chris Wilson wrote:
> > @@ -1713,9 +1711,7 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
> >  
> >  	/* XXX: RCS is the only one to auto invalidate the TLBs? */
> >  	if (engine->id != RCS) {
> > -		ret = engine->emit_flush(req,
> > -					 I915_GEM_GPU_DOMAINS,
> > -					 I915_GEM_GPU_DOMAINS);
> > +		ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);
> 
> Not quite sure why you want to use a combination of flags vs. _BARRIER
> at different places?

Just trying to express the intent of the code. Since this code is
concerned about TLB loading, I though keeping the INVALIDATE explicit
was best. Around the MRI we strictly only need a command stream barrier
and so used BARRIER for documentation (in fact both of these instances
of barriers before the operation can be proven to be not required, as
the required barrier is already provided). After the MRI, we can use a
slightly different barrier than a pure flush/invalidate which might show
a miniscule difference in a purpose built microbenchmark.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 1/2] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

If we rewrite the I915_WRITE_TAIL specialisation for the legacy
ringbuffer as submitting the request onto the ringbuffer, we can unify
the vfunc with both execlists and GuC in the next patch.

v2: Drop the modulus from the I915_WRITE_TAIL as it is currently being
applied in intel_ring_advance() after every command packet, and add a
comment explaining why we need the modulus at all.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1469432687-22756-22-git-send-email-chris@chris-wilson.co.uk
---
 drivers/gpu/drm/i915/i915_gem_request.c |  8 ++----
 drivers/gpu/drm/i915/intel_ringbuffer.c | 51 ++++++++++++++++-----------------
 drivers/gpu/drm/i915/intel_ringbuffer.h | 10 +++++--
 3 files changed, 36 insertions(+), 33 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 606b0b8a5f91..a885905df3bb 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -466,15 +466,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
 	 */
 	request->postfix = ring->tail;
 
-	if (i915.enable_execlists) {
+	if (i915.enable_execlists)
 		ret = engine->emit_request(request);
-	} else {
+	else
 		ret = engine->add_request(request);
-
-		request->tail = ring->tail;
-	}
 	/* Not allowed to fail! */
 	WARN(ret, "emit|add_request failed: %d!\n", ret);
+
 	/* Sanity check that the reserved size was large enough. */
 	ret = ring->tail - request_start;
 	if (ret < 0)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 799a7dc02675..3142085b5cc0 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -58,14 +58,6 @@ void intel_ring_update_space(struct intel_ring *ring)
 					 ring->tail, ring->size);
 }
 
-static void __intel_engine_submit(struct intel_engine_cs *engine)
-{
-	struct intel_ring *ring = engine->buffer;
-
-	ring->tail &= ring->size - 1;
-	engine->write_tail(engine, ring->tail);
-}
-
 static int
 gen2_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 {
@@ -412,13 +404,6 @@ gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
 	return gen8_emit_pipe_control(req, flags, scratch_addr);
 }
 
-static void ring_write_tail(struct intel_engine_cs *engine,
-			    u32 value)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-	I915_WRITE_TAIL(engine, value);
-}
-
 u64 intel_engine_get_active_head(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
@@ -532,7 +517,7 @@ static bool stop_ring(struct intel_engine_cs *engine)
 
 	I915_WRITE_CTL(engine, 0);
 	I915_WRITE_HEAD(engine, 0);
-	engine->write_tail(engine, 0);
+	I915_WRITE_TAIL(engine, 0);
 
 	if (!IS_GEN2(dev_priv)) {
 		(void)I915_READ_CTL(engine);
@@ -1469,7 +1454,10 @@ gen6_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
 	intel_ring_emit(ring, req->fence.seqno);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
-	__intel_engine_submit(engine);
+	intel_ring_advance(ring);
+
+	req->tail = ring->tail;
+	engine->submit_request(req);
 
 	return 0;
 }
@@ -1499,7 +1487,9 @@ gen8_render_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, 0);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
 	intel_ring_emit(ring, MI_NOOP);
-	__intel_engine_submit(engine);
+
+	req->tail = ring->tail;
+	engine->submit_request(req);
 
 	return 0;
 }
@@ -1716,11 +1706,21 @@ i9xx_add_request(struct drm_i915_gem_request *req)
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
 	intel_ring_emit(ring, req->fence.seqno);
 	intel_ring_emit(ring, MI_USER_INTERRUPT);
-	__intel_engine_submit(req->engine);
+	intel_ring_advance(ring);
+
+	req->tail = ring->tail;
+	req->engine->submit_request(req);
 
 	return 0;
 }
 
+static void i9xx_submit_request(struct drm_i915_gem_request *request)
+{
+	struct drm_i915_private *dev_priv = request->i915;
+
+	I915_WRITE_TAIL(request->engine, request->tail);
+}
+
 static void
 gen6_irq_enable(struct intel_engine_cs *engine)
 {
@@ -2479,10 +2479,9 @@ void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
 	rcu_read_unlock();
 }
 
-static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
-				     u32 value)
+static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
+	struct drm_i915_private *dev_priv = request->i915;
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
@@ -2506,8 +2505,8 @@ static void gen6_bsd_ring_write_tail(struct intel_engine_cs *engine,
 		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
 
 	/* Now that the ring is fully powered up, update the tail */
-	I915_WRITE_FW(RING_TAIL(engine->mmio_base), value);
-	POSTING_READ_FW(RING_TAIL(engine->mmio_base));
+	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base), request->tail);
+	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
 
 	/* Let the ring send IDLE messages to the GT again,
 	 * and so let it sleep to conserve power when idle.
@@ -2811,7 +2810,7 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv,
 				      struct intel_engine_cs *engine)
 {
 	engine->init_hw = init_ring_common;
-	engine->write_tail = ring_write_tail;
+	engine->submit_request = i9xx_submit_request;
 
 	engine->add_request = i9xx_add_request;
 	if (INTEL_GEN(dev_priv) >= 6)
@@ -2895,7 +2894,7 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 	if (INTEL_GEN(dev_priv) >= 6) {
 		/* gen6 bsd needs a special wa for tail updates */
 		if (IS_GEN6(dev_priv))
-			engine->write_tail = gen6_bsd_ring_write_tail;
+			engine->submit_request = gen6_bsd_submit_request;
 		engine->emit_flush = gen6_bsd_ring_flush;
 		if (INTEL_GEN(dev_priv) < 8)
 			engine->irq_enable_mask = GT_BSD_USER_INTERRUPT;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 0c3c7185d9ad..14d2ea36fb88 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -204,8 +204,6 @@ struct intel_engine_cs {
 
 	int		(*init_context)(struct drm_i915_gem_request *req);
 
-	void		(*write_tail)(struct intel_engine_cs *engine,
-				      u32 value);
 	int		(*add_request)(struct drm_i915_gem_request *req);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
@@ -296,6 +294,7 @@ struct intel_engine_cs {
 #define I915_DISPATCH_SECURE 0x1
 #define I915_DISPATCH_PINNED 0x2
 #define I915_DISPATCH_RS     0x4
+	void		(*submit_request)(struct drm_i915_gem_request *req);
 
 	/**
 	 * List of objects currently involved in rendering from the
@@ -461,6 +460,13 @@ static inline void intel_ring_emit_reg(struct intel_ring *ring, i915_reg_t reg)
 
 static inline void intel_ring_advance(struct intel_ring *ring)
 {
+	/* The modulus is required so that we avoid writing
+	 * request->tail == ring->size, rather than the expected 0,
+	 * into the RING_TAIL register as that can cause a GPU hang.
+	 * As this is only strictly required for the request->tail,
+	 * and only then as we write the value into hardware, we can
+	 * one day remove the modulus after every command packet.
+	 */
 	ring->tail &= ring->size - 1;
 }
 
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 2/2] drm/i915: Move the modulus for ring emission to the register write

[Chris Wilson]

Space reservation is already safe with respect to the ring->size
modulus, but hardware only expects to see values in the range
0...ring->size-1 (inclusive) and so requires the modulus to prevent us
writing the value ring->size instead of 0. As this is only required for
the register itself, we can defer the modulus to the register update and
not perform it after every command packet. We keep the
intel_ring_advance() around in the code to provide demarcation for the
end-of-packet (with then can be compared against intel_ring_begin() as
the number of dwords emitted must match the reserved space).

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Dave Gordon <david.s.gordon@intel.com>
---
 drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
 drivers/gpu/drm/i915/intel_ringbuffer.c |  6 ++++--
 drivers/gpu/drm/i915/intel_ringbuffer.h | 17 +++++++++++++----
 3 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index bf42a66d6624..824f7efe4e64 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
 	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
 	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
 
-	reg_state[CTX_RING_TAIL+1] = rq->tail;
+	reg_state[CTX_RING_TAIL+1] = intel_ring_offset(rq->ring, rq->tail);
 
 	/* True 32b PPGTT with dynamic page allocation: update PDP
 	 * registers and point the unallocated PDPs to scratch page.
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 3142085b5cc0..21d5e8209400 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1718,7 +1718,8 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request)
 {
 	struct drm_i915_private *dev_priv = request->i915;
 
-	I915_WRITE_TAIL(request->engine, request->tail);
+	I915_WRITE_TAIL(request->engine,
+			intel_ring_offset(request->ring, request->tail));
 }
 
 static void
@@ -2505,7 +2506,8 @@ static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
 		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
 
 	/* Now that the ring is fully powered up, update the tail */
-	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base), request->tail);
+	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
+		      intel_ring_offset(request->ring, request->tail));
 	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
 
 	/* Let the ring send IDLE messages to the GT again,
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 14d2ea36fb88..198b541f9b22 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -460,14 +460,23 @@ static inline void intel_ring_emit_reg(struct intel_ring *ring, i915_reg_t reg)
 
 static inline void intel_ring_advance(struct intel_ring *ring)
 {
+	/* Dummy function.
+	 *
+	 * This serves as a placeholder in the code so that the reader
+	 * can compare against the preceeding intel_ring_begin() and
+	 * check that the number of dwords emitted matches the space
+	 * reserved for the command packet (i.e. the value passed to
+	 * intel_ring_begin()).
+	 */
+}
+
+static inline u32 intel_ring_offset(struct intel_ring *ring, u32 value)
+{
 	/* The modulus is required so that we avoid writing
 	 * request->tail == ring->size, rather than the expected 0,
 	 * into the RING_TAIL register as that can cause a GPU hang.
-	 * As this is only strictly required for the request->tail,
-	 * and only then as we write the value into hardware, we can
-	 * one day remove the modulus after every command packet.
 	 */
-	ring->tail &= ring->size - 1;
+	return value & (ring->size - 1);
 }
 
 int __intel_ring_space(int head, int tail, int size);
-- 
2.8.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Ro.CI.BAT: failure for series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev5)

[Patchwork]

== Series Details ==

Series: series starting with [01/31] drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling() (rev5)
URL   : https://patchwork.freedesktop.org/series/10230/
State : failure

== Summary ==

Applying: drm/i915: Reduce breadcrumb lock coverage for intel_engine_enable_signaling()
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/intel_breadcrumbs.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915: Prefer list_first_entry_or_null
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem_gtt.c
M	drivers/gpu/drm/i915/i915_gem_request.c
M	drivers/gpu/drm/i915/i915_gem_shrinker.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915: Only clear the client pointer when tearing down the file
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915: Only drop the batch-pool's object reference
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_gem_batch_pool.c
Falling back to patching base and 3-way merge...
No changes -- Patch already applied.
Applying: drm/i915/cmdparser: Remove stray intel_engine_cs *ring
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/i915_cmd_parser.c
M	drivers/gpu/drm/i915/i915_drv.h
M	drivers/gpu/drm/i915/i915_gem_execbuffer.c
M	drivers/gpu/drm/i915/intel_engine_cs.c
M	drivers/gpu/drm/i915/intel_lrc.c
M	drivers/gpu/drm/i915/intel_ringbuffer.c
M	drivers/gpu/drm/i915/intel_ringbuffer.h
Falling back to patching base and 3-way merge...
Auto-merging drivers/gpu/drm/i915/i915_cmd_parser.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/i915_cmd_parser.c
error: Failed to merge in the changes.
Patch failed at 0005 drm/i915/cmdparser: Remove stray intel_engine_cs *ring
The copy of the patch that failed is found in: .git/rebase-apply/patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 2/2] drm/i915: Move the modulus for ring emission to the register write

[Joonas Lahtinen]

On to, 2016-07-28 at 10:16 +0100, Chris Wilson wrote:
> Space reservation is already safe with respect to the ring->size
> modulus, but hardware only expects to see values in the range
> 0...ring->size-1 (inclusive) and so requires the modulus to prevent us
> writing the value ring->size instead of 0. As this is only required for
> the register itself, we can defer the modulus to the register update and
> not perform it after every command packet. We keep the
> intel_ring_advance() around in the code to provide demarcation for the
> end-of-packet (with then can be compared against intel_ring_begin() as
> the number of dwords emitted must match the reserved space).
> 
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> Cc: Dave Gordon <david.s.gordon@intel.com>

LGTM,

Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>

> ---
>  drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
>  drivers/gpu/drm/i915/intel_ringbuffer.c |  6 ++++--
>  drivers/gpu/drm/i915/intel_ringbuffer.h | 17 +++++++++++++----
>  3 files changed, 18 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index bf42a66d6624..824f7efe4e64 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
>  	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
>  	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
>  
> -	reg_state[CTX_RING_TAIL+1] = rq->tail;
> +	reg_state[CTX_RING_TAIL+1] = intel_ring_offset(rq->ring, rq->tail);
>  
>  	/* True 32b PPGTT with dynamic page allocation: update PDP
>  	 * registers and point the unallocated PDPs to scratch page.
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 3142085b5cc0..21d5e8209400 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -1718,7 +1718,8 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request)
>  {
>  	struct drm_i915_private *dev_priv = request->i915;
>  
> -	I915_WRITE_TAIL(request->engine, request->tail);
> +	I915_WRITE_TAIL(request->engine,
> +			intel_ring_offset(request->ring, request->tail));
>  }
>  
>  static void
> @@ -2505,7 +2506,8 @@ static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
>  		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
>  
>  	/* Now that the ring is fully powered up, update the tail */
> -	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base), request->tail);
> +	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
> +		      intel_ring_offset(request->ring, request->tail));
>  	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
>  
>  	/* Let the ring send IDLE messages to the GT again,
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 14d2ea36fb88..198b541f9b22 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -460,14 +460,23 @@ static inline void intel_ring_emit_reg(struct intel_ring *ring, i915_reg_t reg)
>  
>  static inline void intel_ring_advance(struct intel_ring *ring)
>  {
> +	/* Dummy function.
> +	 *
> +	 * This serves as a placeholder in the code so that the reader
> +	 * can compare against the preceeding intel_ring_begin() and
> +	 * check that the number of dwords emitted matches the space
> +	 * reserved for the command packet (i.e. the value passed to
> +	 * intel_ring_begin()).
> +	 */
> +}
> +
> +static inline u32 intel_ring_offset(struct intel_ring *ring, u32 value)
> +{
>  	/* The modulus is required so that we avoid writing
>  	 * request->tail == ring->size, rather than the expected 0,
>  	 * into the RING_TAIL register as that can cause a GPU hang.
> -	 * As this is only strictly required for the request->tail,
> -	 * and only then as we write the value into hardware, we can
> -	 * one day remove the modulus after every command packet.
>  	 */
> -	ring->tail &= ring->size - 1;
> +	return value & (ring->size - 1);
>  }
>  
>  int __intel_ring_space(int head, int tail, int size);
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Reduce engine->emit_flush() to a single mode parameter

[Joonas Lahtinen]

On to, 2016-07-28 at 09:37 +0100, Chris Wilson wrote:
> On Thu, Jul 28, 2016 at 10:11:43AM +0300, Joonas Lahtinen wrote:
> > 
> > On ke, 2016-07-27 at 11:53 +0100, Chris Wilson wrote:
> > > 
> > > @@ -1713,9 +1711,7 @@ static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
> > >  
> > >  	/* XXX: RCS is the only one to auto invalidate the TLBs? */
> > >  	if (engine->id != RCS) {
> > > -		ret = engine->emit_flush(req,
> > > -					 I915_GEM_GPU_DOMAINS,
> > > -					 I915_GEM_GPU_DOMAINS);
> > > +		ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);
> > Not quite sure why you want to use a combination of flags vs. _BARRIER
> > at different places?
> Just trying to express the intent of the code. Since this code is
> concerned about TLB loading, I though keeping the INVALIDATE explicit
> was best. Around the MRI we strictly only need a command stream barrier
> and so used BARRIER for documentation (in fact both of these instances
> of barriers before the operation can be proven to be not required, as
> the required barrier is already provided). After the MRI, we can use a
> slightly different barrier than a pure flush/invalidate which might show
> a miniscule difference in a purpose built microbenchmark.

I do not have a strong opinion, guess it is fine either way.

Regards, Joonas

> -Chris
> 
-- 
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Dave Gordon]

On 27/07/16 16:28, Chris Wilson wrote:
> On Wed, Jul 27, 2016 at 12:08:35PM +0100, Dave Gordon wrote:
>> On 25/07/16 10:18, Joonas Lahtinen wrote:
>>> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
>>>> If is simpler and leads to more readable code through the callstack if
>>>> the allocation returns the allocated struct through the return value.
>>>>
>>>> The importance of this is that it no longer looks like we accidentally
>>>> allocate requests as side-effect of calling .
>>>
>>> Dave seems to have expressed to wish to review this around January, so
>>> CC'ing him here.
>>>
>> >From me,
>>>
>>> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>>>
>>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>>> ---
>>>> drivers/gpu/drm/i915/i915_drv.h            |  3 +-
>>>> drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
>>>> drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
>>>> drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
>>>> drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
>>>> drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
>>>> drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
>>>> drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
>>>> 8 files changed, 79 insertions(+), 140 deletions(-)
>>
>> The actual code looks fine, but the patch amalgamates several
>> different changes, only one  of which is mentioned in the
>> description above.
>>
>> 1. Change the signature of i915_gem_object_sync() and its internal
>>    subfunction __i915_gem_object_sync() to always require a request
>>    as input, with corresponding rework of the callsites. This enables
>>    the removal of the internal call to i915_gem_request_alloc().
>>
>> 2. Change API of i915_gem_request_alloc() not to allow NULL ctx,
>>    which involves changing various of the remaining callsites.
>
> Reverts the unwanted the change.

No, 2 and 3 together do NOT revert the change 268270883 because that's 
what originally changed the API of i915_gem_request_alloc() to return a 
pointer directly, rather than via an output parameter. Which everyone 
agrees is a good idea :)

Allowing NULL here was a useful step in the elimination of the per-ring 
default context (which was after all your idea, IIRC). Now that there 
aren't so many callsites it's not so ugly to have those callers mention 
the (now-unified) kernel context.

>> 3. Flatten __i915_gem_request_alloc() into i915_gem_request_alloc().
>>
>> The description really only covers (3), which is the simplest and
>> most mechanical of the changes. So preferably two or three separate
>> patches, in whatever order makes the most sense, with a bit more
>> description of each.
>
> It mentions 1 explicitly which was the reason why the API was originally
> intended to be this.
> -Chris

Strange, I can't see any *explicit* mention of i915_gem_object_sync(), 
only some vague allusion to "certain functions".

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH] drm/i915: Reduce engine->emit_flush() to a single mode parameter

[Dave Gordon]

On 27/07/16 11:53, Chris Wilson wrote:
> Rather than passing a complete set of GPU cache domains for either
> invalidation or for flushing, or even both, just pass a single parameter
> to the engine->emit_flush to determine the required operations.
>
> engine->emit_flush(GPU, 0) -> engine->emit_flush(EMIT_INVALIDATE)
> engine->emit_flush(0, GPU) -> engine->emit_flush(EMIT_FLUSH)
> engine->emit_flush(GPU, GPU) -> engine->emit_flush(EMIT_FLUSH | EMIT_INVALIDATE)
>
> This allows us to extend the behaviour easily in future, for example if
> we want just a command barrier without the overhead of flushing.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Dave Gordon <david.s.gordon@intel.com>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/i915_gem_context.c    |  2 +-
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
>  drivers/gpu/drm/i915/i915_gem_gtt.c        | 10 ++----
>  drivers/gpu/drm/i915/i915_gem_request.c    |  2 +-
>  drivers/gpu/drm/i915/intel_lrc.c           | 23 +++++-------
>  drivers/gpu/drm/i915/intel_ringbuffer.c    | 57 +++++++++++-------------------
>  drivers/gpu/drm/i915/intel_ringbuffer.h    |  6 ++--
>  7 files changed, 38 insertions(+), 64 deletions(-)

[snip]

> diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
> index 39fa9eb10514..671b1cab5e54 100644
> --- a/drivers/gpu/drm/i915/i915_gem_gtt.c
> +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
> @@ -1666,8 +1666,7 @@ static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
>  	int ret;
>
>  	/* NB: TLBs must be flushed and invalidated before a switch */
> -	ret = engine->emit_flush(req,
> -				 I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
> +	ret = engine->emit_flush(req, EMIT_INVALIDATE | EMIT_FLUSH);

The example in the commit message says "EMIT_FLUSH | EMIT_INVALIDATE"
which I think looks much nicer. Flush *after* invalidate would be pretty 
meaningless!

For even more syntactic sugar, we could choose "FLUSH + INVALIDATE" as + 
and | are equivalent for disjoint bitfields.

>  int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
> @@ -998,9 +998,7 @@ static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
>  	if (w->count == 0)
>  		return 0;
>
> -	ret = req->engine->emit_flush(req,
> -				      I915_GEM_GPU_DOMAINS,
> -				      I915_GEM_GPU_DOMAINS);
> +	ret = req->engine->emit_flush(req, EMIT_BARRIER);
>  	if (ret)
>  		return ret;

Distinguishing flush-and-invalidate from BARRIER seems like a good idea, 
because here we really don't want to flush or invalidate the GPU's 
caches; we really only want some sort of synchronisation of memory 
activity before changing the register. Therefore ...

[snip]

> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 00723401f98c..76d0495943c3 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -292,8 +292,10 @@ struct intel_engine_cs {
>  	u32 ctx_desc_template;
>  	int		(*emit_request)(struct drm_i915_gem_request *request);
>  	int		(*emit_flush)(struct drm_i915_gem_request *request,
> -				      u32 invalidate_domains,
> -				      u32 flush_domains);
> +				      u32 mode);
> +#define EMIT_INVALIDATE	BIT(0)
> +#define EMIT_FLUSH	BIT(1)
> +#define EMIT_BARRIER	(EMIT_INVALIDATE | EMIT_FLUSH)

... I think these should be three (or maybe four) distinct operations, 
and collapsing BARRIER to FLUSH+INVALIDATE if the h/w can't do a simple 
barrier should be left to the gen-specific code. Also what about 
allowing for a FLUSH or INVALIDATE without BARRIER semantics? Do we care 
about whether the command streamer stalls before/after these operations?

So maybe start with

#define	EMIT_CS_STALL	BIT(0)	/* Stall CS before operation	*/
#define	EMIT_WRITEBACK	BIT(1)	/* Ensure caches written back	*/
#define	EMIT_DROP	BIT(2)	/* Drop contents of cache(s)	*/
#define	EMIT_POSTSYNC	BIT(3)	/* Wait until complete		*/

and then

#define	EMIT_BARRIER	(EMIT_CS_STALL+EMIT_POSTSYNC)
#define	EMIT_FLUSH	(EMIT_CS_STALL+EMIT_WRITEBACK)
#define	EMIT_INVALIDATE	(EMIT_DROP+EMIT_POSTSYNC)
#define	EMIT_SWITCH	(EMIT_FLUSH+EMIT_INVALIDATE)

or whatever combinations of basic operations are actually useful.

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Dave Gordon]

On 27/07/16 13:29, Chris Wilson wrote:
> On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
>> On 25/07/16 08:44, Chris Wilson wrote:
>>> If we rewrite the I915_WRITE_TAIL specialisation for the legacy
>>> ringbuffer as submitting the request onto the ringbuffer, we can unify
>>> the vfunc with both execlists and GuC in the next patch.
>>>
>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>>> ---
>>> drivers/gpu/drm/i915/i915_gem_request.c |  8 ++---
>>> drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
>>> drivers/gpu/drm/i915/intel_ringbuffer.c | 53 +++++++++++++++++----------------
>>> drivers/gpu/drm/i915/intel_ringbuffer.h |  3 +-
>>> 4 files changed, 32 insertions(+), 34 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
>>> index 1c185e293bf0..8814e9c5266b 100644
>>> --- a/drivers/gpu/drm/i915/i915_gem_request.c
>>> +++ b/drivers/gpu/drm/i915/i915_gem_request.c
>>> @@ -467,15 +467,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
>>> 	 */
>>> 	request->postfix = ring->tail;
>>>
>>> -	if (i915.enable_execlists) {
>>> +	if (i915.enable_execlists)
>>> 		ret = engine->emit_request(request);
>>> -	} else {
>>> +	else
>>> 		ret = engine->add_request(request);
>>> -
>>> -		request->tail = ring->tail;
>>> -	}
>>> 	/* Not allowed to fail! */
>>> 	WARN(ret, "emit|add_request failed: %d!\n", ret);
>>> +
>>> 	/* Sanity check that the reserved size was large enough. */
>>> 	ret = ring->tail - request_start;
>>> 	if (ret < 0)
>>> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
>>> index 567d94de3300..250edb2bcef7 100644
>>> --- a/drivers/gpu/drm/i915/intel_lrc.c
>>> +++ b/drivers/gpu/drm/i915/intel_lrc.c
>>> @@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
>>> 	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
>>> 	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
>>>
>>> -	reg_state[CTX_RING_TAIL+1] = rq->tail;
>>> +	reg_state[CTX_RING_TAIL+1] = rq->tail % (rq->ring->size - 1);
>>
>> mod ringsize-1 ?
>>
>> Surely tail % ringsize, or tail & (ringsize-1).
>>
>> But it's redundant anyway, rq->tail cannot exceed ring->size,
>> so the original code was correct.
>
> No, rq->tail can be equal to ring->size which leads to a GPU hang.
> (Observed on the older gen at least, I'd rather have the same paranoia
> here.)
> -Chris

Even if it's not redundant, it's still the wrong number. The code above 
would result in tail (==size) being converted to 1 rather than 0.

If it's a % operation, it should be ringsize not ringsize-1. Or convert 
to an & operation with ringsize-1.

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 21/31] drm/i915: Convert engine->write_tail to operate on a request

[Chris Wilson]

On Thu, Jul 28, 2016 at 04:05:28PM +0100, Dave Gordon wrote:
> On 27/07/16 13:29, Chris Wilson wrote:
> >On Wed, Jul 27, 2016 at 12:53:25PM +0100, Dave Gordon wrote:
> >>On 25/07/16 08:44, Chris Wilson wrote:
> >>>If we rewrite the I915_WRITE_TAIL specialisation for the legacy
> >>>ringbuffer as submitting the request onto the ringbuffer, we can unify
> >>>the vfunc with both execlists and GuC in the next patch.
> >>>
> >>>Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >>>Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> >>>---
> >>>drivers/gpu/drm/i915/i915_gem_request.c |  8 ++---
> >>>drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
> >>>drivers/gpu/drm/i915/intel_ringbuffer.c | 53 +++++++++++++++++----------------
> >>>drivers/gpu/drm/i915/intel_ringbuffer.h |  3 +-
> >>>4 files changed, 32 insertions(+), 34 deletions(-)
> >>>
> >>>diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
> >>>index 1c185e293bf0..8814e9c5266b 100644
> >>>--- a/drivers/gpu/drm/i915/i915_gem_request.c
> >>>+++ b/drivers/gpu/drm/i915/i915_gem_request.c
> >>>@@ -467,15 +467,13 @@ void __i915_add_request(struct drm_i915_gem_request *request,
> >>>	 */
> >>>	request->postfix = ring->tail;
> >>>
> >>>-	if (i915.enable_execlists) {
> >>>+	if (i915.enable_execlists)
> >>>		ret = engine->emit_request(request);
> >>>-	} else {
> >>>+	else
> >>>		ret = engine->add_request(request);
> >>>-
> >>>-		request->tail = ring->tail;
> >>>-	}
> >>>	/* Not allowed to fail! */
> >>>	WARN(ret, "emit|add_request failed: %d!\n", ret);
> >>>+
> >>>	/* Sanity check that the reserved size was large enough. */
> >>>	ret = ring->tail - request_start;
> >>>	if (ret < 0)
> >>>diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> >>>index 567d94de3300..250edb2bcef7 100644
> >>>--- a/drivers/gpu/drm/i915/intel_lrc.c
> >>>+++ b/drivers/gpu/drm/i915/intel_lrc.c
> >>>@@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
> >>>	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
> >>>	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
> >>>
> >>>-	reg_state[CTX_RING_TAIL+1] = rq->tail;
> >>>+	reg_state[CTX_RING_TAIL+1] = rq->tail % (rq->ring->size - 1);
> >>
> >>mod ringsize-1 ?
> >>
> >>Surely tail % ringsize, or tail & (ringsize-1).
> >>
> >>But it's redundant anyway, rq->tail cannot exceed ring->size,
> >>so the original code was correct.
> >
> >No, rq->tail can be equal to ring->size which leads to a GPU hang.
> >(Observed on the older gen at least, I'd rather have the same paranoia
> >here.)
> >-Chris
> 
> Even if it's not redundant, it's still the wrong number. The code
> above would result in tail (==size) being converted to 1 rather than
> 0.
> 
> If it's a % operation, it should be ringsize not ringsize-1. Or
> convert to an & operation with ringsize-1.

It was just meant to be & (ring->size-1).
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Chris Wilson]

On Thu, Jul 28, 2016 at 01:48:32PM +0100, Dave Gordon wrote:
> On 27/07/16 16:28, Chris Wilson wrote:
> >On Wed, Jul 27, 2016 at 12:08:35PM +0100, Dave Gordon wrote:
> >>On 25/07/16 10:18, Joonas Lahtinen wrote:
> >>>On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
> >>>>If is simpler and leads to more readable code through the callstack if
> >>>>the allocation returns the allocated struct through the return value.
> >>>>
> >>>>The importance of this is that it no longer looks like we accidentally
> >>>>allocate requests as side-effect of calling .
> >>>
> >>>Dave seems to have expressed to wish to review this around January, so
> >>>CC'ing him here.
> >>>
> >>>From me,
> >>>
> >>>Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> >>>
> >>>>Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >>>>---
> >>>>drivers/gpu/drm/i915/i915_drv.h            |  3 +-
> >>>>drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
> >>>>drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
> >>>>drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
> >>>>drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
> >>>>drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
> >>>>drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
> >>>>drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
> >>>>8 files changed, 79 insertions(+), 140 deletions(-)
> >>
> >>The actual code looks fine, but the patch amalgamates several
> >>different changes, only one  of which is mentioned in the
> >>description above.
> >>
> >>1. Change the signature of i915_gem_object_sync() and its internal
> >>   subfunction __i915_gem_object_sync() to always require a request
> >>   as input, with corresponding rework of the callsites. This enables
> >>   the removal of the internal call to i915_gem_request_alloc().
> >>
> >>2. Change API of i915_gem_request_alloc() not to allow NULL ctx,
> >>   which involves changing various of the remaining callsites.
> >
> >Reverts the unwanted the change.
> 
> No, 2 and 3 together do NOT revert the change 268270883 because
> that's what originally changed the API of i915_gem_request_alloc()
> to return a pointer directly, rather than via an output parameter.
> Which everyone agrees is a good idea :)
> 
> Allowing NULL here was a useful step in the elimination of the
> per-ring default context (which was after all your idea, IIRC). Now
> that there aren't so many callsites it's not so ugly to have those
> callers mention the (now-unified) kernel context.

I'm referring to the change that took it from the API here to what was
merged, and so reverting back to what I originally wrote.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 2/2] drm/i915: Move the modulus for ring emission to the register write

[Dave Gordon]

On 28/07/16 10:16, Chris Wilson wrote:
> Space reservation is already safe with respect to the ring->size
> modulus, but hardware only expects to see values in the range
> 0...ring->size-1 (inclusive) and so requires the modulus to prevent us
> writing the value ring->size instead of 0. As this is only required for
> the register itself, we can defer the modulus to the register update and
> not perform it after every command packet. We keep the
> intel_ring_advance() around in the code to provide demarcation for the
> end-of-packet (with then can be compared against intel_ring_begin() as
> the number of dwords emitted must match the reserved space).
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> Cc: Dave Gordon <david.s.gordon@intel.com>
> ---
>  drivers/gpu/drm/i915/intel_lrc.c        |  2 +-
>  drivers/gpu/drm/i915/intel_ringbuffer.c |  6 ++++--
>  drivers/gpu/drm/i915/intel_ringbuffer.h | 17 +++++++++++++----
>  3 files changed, 18 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index bf42a66d6624..824f7efe4e64 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -373,7 +373,7 @@ static void execlists_update_context(struct drm_i915_gem_request *rq)
>  	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
>  	uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
>
> -	reg_state[CTX_RING_TAIL+1] = rq->tail;
> +	reg_state[CTX_RING_TAIL+1] = intel_ring_offset(rq->ring, rq->tail);
>
>  	/* True 32b PPGTT with dynamic page allocation: update PDP
>  	 * registers and point the unallocated PDPs to scratch page.
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 3142085b5cc0..21d5e8209400 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -1718,7 +1718,8 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request)
>  {
>  	struct drm_i915_private *dev_priv = request->i915;
>
> -	I915_WRITE_TAIL(request->engine, request->tail);
> +	I915_WRITE_TAIL(request->engine,
> +			intel_ring_offset(request->ring, request->tail));
>  }
>
>  static void
> @@ -2505,7 +2506,8 @@ static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
>  		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
>
>  	/* Now that the ring is fully powered up, update the tail */
> -	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base), request->tail);
> +	I915_WRITE_FW(RING_TAIL(request->engine->mmio_base),
> +		      intel_ring_offset(request->ring, request->tail));
>  	POSTING_READ_FW(RING_TAIL(request->engine->mmio_base));
>
>  	/* Let the ring send IDLE messages to the GT again,
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
> index 14d2ea36fb88..198b541f9b22 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.h
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
> @@ -460,14 +460,23 @@ static inline void intel_ring_emit_reg(struct intel_ring *ring, i915_reg_t reg)
>
>  static inline void intel_ring_advance(struct intel_ring *ring)
>  {
> +	/* Dummy function.
> +	 *
> +	 * This serves as a placeholder in the code so that the reader
> +	 * can compare against the preceeding intel_ring_begin() and
> +	 * check that the number of dwords emitted matches the space
> +	 * reserved for the command packet (i.e. the value passed to
> +	 * intel_ring_begin()).
> +	 */
> +}
> +
> +static inline u32 intel_ring_offset(struct intel_ring *ring, u32 value)
> +{
>  	/* The modulus is required so that we avoid writing
>  	 * request->tail == ring->size, rather than the expected 0,
>  	 * into the RING_TAIL register as that can cause a GPU hang.
> -	 * As this is only strictly required for the request->tail,
> -	 * and only then as we write the value into hardware, we can
> -	 * one day remove the modulus after every command packet.
>  	 */
> -	ring->tail &= ring->size - 1;
> +	return value & (ring->size - 1);
>  }

Hmmm ... better, but the comment refers to a "modulus" operation while 
the code uses an "&". With the longer commentary at the top of the 
patch, maybe this could now just say something like

/* ensure value (TAIL) is strictly less than ring->size */

BTW, a completely different way to avoid this problem would just be to 
ensure the packet never finished at the last word of the ring. Setting 
effective_size = size-1 would do just that!

.Dave.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 18/31] drm/i915: Simplify request_alloc by returning the allocated request

[Dave Gordon]

On 28/07/16 16:10, Chris Wilson wrote:
> On Thu, Jul 28, 2016 at 01:48:32PM +0100, Dave Gordon wrote:
>> On 27/07/16 16:28, Chris Wilson wrote:
>>> On Wed, Jul 27, 2016 at 12:08:35PM +0100, Dave Gordon wrote:
>>>> On 25/07/16 10:18, Joonas Lahtinen wrote:
>>>>> On ma, 2016-07-25 at 08:44 +0100, Chris Wilson wrote:
>>>>>> If is simpler and leads to more readable code through the callstack if
>>>>>> the allocation returns the allocated struct through the return value.
>>>>>>
>>>>>> The importance of this is that it no longer looks like we accidentally
>>>>>> allocate requests as side-effect of calling .
>>>>>
>>>>> Dave seems to have expressed to wish to review this around January, so
>>>>> CC'ing him here.
>>>>>
>>>> >From me,
>>>>>
>>>>> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>>>>>
>>>>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>>>>> ---
>>>>>> drivers/gpu/drm/i915/i915_drv.h            |  3 +-
>>>>>> drivers/gpu/drm/i915/i915_gem.c            | 75 ++++++++----------------------
>>>>>> drivers/gpu/drm/i915/i915_gem_execbuffer.c | 12 ++---
>>>>>> drivers/gpu/drm/i915/i915_gem_request.c    | 58 ++++++++---------------
>>>>>> drivers/gpu/drm/i915/i915_trace.h          | 13 +++---
>>>>>> drivers/gpu/drm/i915/intel_display.c       | 36 ++++++--------
>>>>>> drivers/gpu/drm/i915/intel_lrc.c           |  2 +-
>>>>>> drivers/gpu/drm/i915/intel_overlay.c       | 20 ++++----
>>>>>> 8 files changed, 79 insertions(+), 140 deletions(-)
>>>>
>>>> The actual code looks fine, but the patch amalgamates several
>>>> different changes, only one  of which is mentioned in the
>>>> description above.
>>>>
>>>> 1. Change the signature of i915_gem_object_sync() and its internal
>>>>   subfunction __i915_gem_object_sync() to always require a request
>>>>   as input, with corresponding rework of the callsites. This enables
>>>>   the removal of the internal call to i915_gem_request_alloc().
>>>>
>>>> 2. Change API of i915_gem_request_alloc() not to allow NULL ctx,
>>>>   which involves changing various of the remaining callsites.
>>>
>>> Reverts the unwanted the change.
>>
>> No, 2 and 3 together do NOT revert the change 268270883 because
>> that's what originally changed the API of i915_gem_request_alloc()
>> to return a pointer directly, rather than via an output parameter.
>> Which everyone agrees is a good idea :)
>>
>> Allowing NULL here was a useful step in the elimination of the
>> per-ring default context (which was after all your idea, IIRC). Now
>> that there aren't so many callsites it's not so ugly to have those
>> callers mention the (now-unified) kernel context.
>
> I'm referring to the change that took it from the API here to what was
> merged, and so reverting back to what I originally wrote.
> -Chris

Whatever. Item 1 still needs to be a separate patch from 2&|3.

.Dave.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 25/31] drm/i915: Reuse legacy breadcrumbs + tail emission

[Dave Gordon]

On 25/07/16 08:44, Chris Wilson wrote:
> As GEN6+ is now a simple variant on the basic breadcrumbs + tail write,
> reuse the common code.
>
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> ---
>  drivers/gpu/drm/i915/intel_ringbuffer.c | 74 +++++++++++++--------------------
>  1 file changed, 30 insertions(+), 44 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 68df689a9d1e..a42d878009e6 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -1428,26 +1428,19 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req)
>  	return 0;
>  }
>
> -/**
> - * gen6_emit_request - Update the semaphore mailbox registers
> - *
> - * @request - request to write to the ring
> - *
> - * Update the mailbox registers in the *other* rings with the current seqno.
> - * This acts like a signal in the canonical semaphore.
> - */
> -static int gen6_emit_request(struct drm_i915_gem_request *req)
> +static void i9xx_submit_request(struct drm_i915_gem_request *request)
> +{
> +	struct drm_i915_private *dev_priv = request->i915;
> +
> +	I915_WRITE_TAIL(request->engine,
> +			request->tail % (request->ring->size - 1));
> +}

wrong mod-thing as previously noted - I don't know whether this is 
already fixed up elsewhere.

.Dave.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 25/31] drm/i915: Reuse legacy breadcrumbs + tail emission

[Chris Wilson]

On Thu, Jul 28, 2016 at 04:23:42PM +0100, Dave Gordon wrote:
> On 25/07/16 08:44, Chris Wilson wrote:
> >As GEN6+ is now a simple variant on the basic breadcrumbs + tail write,
> >reuse the common code.
> >
> >Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> >Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
> >---
> > drivers/gpu/drm/i915/intel_ringbuffer.c | 74 +++++++++++++--------------------
> > 1 file changed, 30 insertions(+), 44 deletions(-)
> >
> >diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> >index 68df689a9d1e..a42d878009e6 100644
> >--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> >+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> >@@ -1428,26 +1428,19 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req)
> > 	return 0;
> > }
> >
> >-/**
> >- * gen6_emit_request - Update the semaphore mailbox registers
> >- *
> >- * @request - request to write to the ring
> >- *
> >- * Update the mailbox registers in the *other* rings with the current seqno.
> >- * This acts like a signal in the canonical semaphore.
> >- */
> >-static int gen6_emit_request(struct drm_i915_gem_request *req)
> >+static void i9xx_submit_request(struct drm_i915_gem_request *request)
> >+{
> >+	struct drm_i915_private *dev_priv = request->i915;
> >+
> >+	I915_WRITE_TAIL(request->engine,
> >+			request->tail % (request->ring->size - 1));
> >+}
> 
> wrong mod-thing as previously noted - I don't know whether this is
> already fixed up elsewhere.

You previously replied to altered patches with the recent changes undone
in favour of introducing intel_ring_offset().
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [PATCH 25/31] drm/i915: Reuse legacy breadcrumbs + tail emission

[Dave Gordon]

On 28/07/16 16:29, Chris Wilson wrote:
> On Thu, Jul 28, 2016 at 04:23:42PM +0100, Dave Gordon wrote:
>> On 25/07/16 08:44, Chris Wilson wrote:
>>> As GEN6+ is now a simple variant on the basic breadcrumbs + tail write,
>>> reuse the common code.
>>>
>>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>>> Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>>> ---
>>> drivers/gpu/drm/i915/intel_ringbuffer.c | 74 +++++++++++++--------------------
>>> 1 file changed, 30 insertions(+), 44 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
>>> index 68df689a9d1e..a42d878009e6 100644
>>> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
>>> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
>>> @@ -1428,26 +1428,19 @@ static int gen6_signal(struct drm_i915_gem_request *signaller_req)
>>> 	return 0;
>>> }
>>>
>>> -/**
>>> - * gen6_emit_request - Update the semaphore mailbox registers
>>> - *
>>> - * @request - request to write to the ring
>>> - *
>>> - * Update the mailbox registers in the *other* rings with the current seqno.
>>> - * This acts like a signal in the canonical semaphore.
>>> - */
>>> -static int gen6_emit_request(struct drm_i915_gem_request *req)
>>> +static void i9xx_submit_request(struct drm_i915_gem_request *request)
>>> +{
>>> +	struct drm_i915_private *dev_priv = request->i915;
>>> +
>>> +	I915_WRITE_TAIL(request->engine,
>>> +			request->tail % (request->ring->size - 1));
>>> +}
>>
>> wrong mod-thing as previously noted - I don't know whether this is
>> already fixed up elsewhere.
>
> You previously replied to altered patches with the recent changes undone
> in favour of introducing intel_ring_offset().
> -Chris

Fine, I just wasn't sure what got superseded in which string-of-patches 
(as there seem to be quite a lot running in parallel here) and didn't 
want this instance to get missed (as it had already got an R-B).

.Dave.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx