[Intel-gfx] [PATCH] drm/i915: Tracking waiters for execution wrt hangcheck

[Chris Wilson]

Hangcheck wants to turn itself off when no one is listening (to save
needless wakeups). However, we now may have waiters that are not waiting
for requests already on the GPU, but waiting for requests deep in the
queue for execution, and do not have an irq alloted. Therefore, inside
hangcheck we can not simply just check if any waiter has an irq
assigned, but can look at whether anyone is listening for hangcheck
signals.

Fixes: 4680816be336 ("drm/i915: Wait first for submission, before waiting for request completion")
Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala at linux.intel.com>
---
 drivers/gpu/drm/i915/intel_hangcheck.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_hangcheck.c b/drivers/gpu/drm/i915/intel_hangcheck.c
index 31f01d64c021..34db72e635fc 100644
--- a/drivers/gpu/drm/i915/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/intel_hangcheck.c
@@ -411,7 +411,6 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	unsigned int hung = 0, stuck = 0;
-	int busy_count = 0;
 
 	if (!i915_modparams.enable_hangcheck)
 		return;
@@ -429,7 +428,6 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 	intel_uncore_arm_unclaimed_mmio_detection(dev_priv);
 
 	for_each_engine(engine, dev_priv, id) {
-		const bool busy = intel_engine_has_waiter(engine);
 		struct intel_engine_hangcheck hc;
 
 		semaphore_clear_deadlocks(dev_priv);
@@ -443,15 +441,13 @@ static void i915_hangcheck_elapsed(struct work_struct *work)
 			if (hc.action != ENGINE_DEAD)
 				stuck |= intel_engine_flag(engine);
 		}
-
-		busy_count += busy;
 	}
 
 	if (hung)
 		hangcheck_declare_hang(dev_priv, hung, stuck);
 
 	/* Reset timer in case GPU hangs without another request being added */
-	if (busy_count)
+	if (waitqueue_active(&dev_priv->gpu_error.wait_queue))
 		i915_queue_hangcheck(dev_priv);
 }
 
-- 
2.15.1