[Intel-gfx] [PATCH 1/2] drm/i915: ban badly behaving contexts

[Mika Kuoppala]

Now when we have mechanism in place to track which context
was guilty of hanging the gpu, it is possible to punish
for bad behaviour.

If context has recently submitted a faulty batchbuffers guilty of
gpu hang and submits another batch which hangs gpu in quick
succession, ban it permanently. If ctx is banned, no more
batchbuffers will be queued for execution.

There is no need for global wedge machinery anymore and
it would be unwise to wedge the whole gpu if we have multiple
hanging batches queued for execution. Instead just ban
the guilty ones and carry on.

v2: Store guilty ban status bool in gpu_error instead of pointers
    that might become danling before hang is declared.

v3: Use return value for banned status instead of stashing state
    into gpu_error (Chris Wilson)

v4: - rebase on top of fixed hang stats api
    - add define for ban period
    - rename commit and improve commit msg

v5: - rely context banning instead of wedging the gpu
    - beautification and fix for ban calculation (Chris)

Signed-off-by: Mika Kuoppala <mika.kuoppala at intel.com>
---
 drivers/gpu/drm/i915/i915_drv.c            |   29 ++++++++++++----------------
 drivers/gpu/drm/i915/i915_drv.h            |   11 +++++++++--
 drivers/gpu/drm/i915/i915_gem.c            |   22 +++++++++++++++++++--
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |   12 ++++++++++++
 4 files changed, 53 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index eb56f68..a3beef3 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -719,24 +719,19 @@ int i915_reset(struct drm_device *dev)
 
 	simulated = dev_priv->gpu_error.stop_rings != 0;
 
-	if (!simulated && get_seconds() - dev_priv->gpu_error.last_reset < 5) {
-		DRM_ERROR("GPU hanging too fast, declaring wedged!\n");
-		ret = -ENODEV;
-	} else {
-		ret = intel_gpu_reset(dev);
-
-		/* Also reset the gpu hangman. */
-		if (simulated) {
-			DRM_INFO("Simulated gpu hang, resetting stop_rings\n");
-			dev_priv->gpu_error.stop_rings = 0;
-			if (ret == -ENODEV) {
-				DRM_ERROR("Reset not implemented, but ignoring "
-					  "error for simulated gpu hangs\n");
-				ret = 0;
-			}
-		} else
-			dev_priv->gpu_error.last_reset = get_seconds();
+	ret = intel_gpu_reset(dev);
+
+	/* Also reset the gpu hangman. */
+	if (simulated) {
+		DRM_INFO("Simulated gpu hang, resetting stop_rings\n");
+		dev_priv->gpu_error.stop_rings = 0;
+		if (ret == -ENODEV) {
+			DRM_ERROR("Reset not implemented, but ignoring "
+				  "error for simulated gpu hangs\n");
+			ret = 0;
+		}
 	}
+
 	if (ret) {
 		DRM_ERROR("Failed to reset chip.\n");
 		mutex_unlock(&dev->struct_mutex);
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 4da1f86..2c5f3bc 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -583,6 +583,12 @@ struct i915_ctx_hang_stats {
 
 	/* This context had batch active when hang was declared */
 	unsigned batch_active;
+
+	/* Time when this context was last blamed for a GPU reset */
+	unsigned long guilty_ts;
+
+	/* This context is banned to submit more work */
+	bool banned;
 };
 
 /* This must match up with the value previously used for execbuf2.rsvd1. */
@@ -984,6 +990,9 @@ struct i915_gpu_error {
 	/* For hangcheck timer */
 #define DRM_I915_HANGCHECK_PERIOD 1500 /* in ms */
 #define DRM_I915_HANGCHECK_JIFFIES msecs_to_jiffies(DRM_I915_HANGCHECK_PERIOD)
+	/* Hang gpu twice in this window and your context gets banned */
+#define DRM_I915_CTX_BAN_PERIOD DIV_ROUND_UP(8*DRM_I915_HANGCHECK_PERIOD, 1000)
+
 	struct timer_list hangcheck_timer;
 
 	/* For reset and error_state handling. */
@@ -992,8 +1001,6 @@ struct i915_gpu_error {
 	struct drm_i915_error_state *first_error;
 	struct work_struct work;
 
-	unsigned long last_reset;
-
 	/**
 	 * State variable and reset counter controlling the reset flow
 	 *
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 7392f64..754f875 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2201,6 +2201,21 @@ static bool i915_request_guilty(struct drm_i915_gem_request *request,
 	return false;
 }
 
+static bool i915_context_is_banned(const struct i915_ctx_hang_stats *hs)
+{
+	const unsigned long elapsed = get_seconds() - hs->guilty_ts;
+
+	if (hs->banned)
+		return true;
+
+	if (elapsed <= DRM_I915_CTX_BAN_PERIOD) {
+		DRM_ERROR("context hanging too fast, declaring banned!\n");
+		return true;
+	}
+
+	return false;
+}
+
 static void i915_set_reset_status(struct intel_ring_buffer *ring,
 				  struct drm_i915_gem_request *request,
 				  u32 acthd)
@@ -2237,10 +2252,13 @@ static void i915_set_reset_status(struct intel_ring_buffer *ring,
 		hs = &request->file_priv->hang_stats;
 
 	if (hs) {
-		if (guilty)
+		if (guilty) {
+			hs->banned = i915_context_is_banned(hs);
 			hs->batch_active++;
-		else
+			hs->guilty_ts = get_seconds();
+		} else {
 			hs->batch_pending++;
+		}
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index ea4bacb..426021a 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -926,6 +926,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *batch_obj;
 	struct drm_clip_rect *cliprects = NULL;
 	struct intel_ring_buffer *ring;
+	struct i915_ctx_hang_stats *hs;
 	u32 ctx_id = i915_execbuffer2_get_context_id(*args);
 	u32 exec_start, exec_len;
 	u32 mask, flags;
@@ -1115,6 +1116,17 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (ret)
 		goto err;
 
+	hs = i915_gem_context_get_hang_stats(dev, file, ctx_id);
+	if (IS_ERR(hs)) {
+		ret = PTR_ERR(hs);
+		goto err;
+	}
+
+	if (hs->banned) {
+		ret = -EIO;
+		goto err;
+	}
+
 	ret = i915_switch_context(ring, file, ctx_id);
 	if (ret)
 		goto err;
-- 
1.7.9.5