drm/i915: Fix __wait_seqno to use true infinite timeouts

When we switched to always using a timeout in conjunction with
wait_seqno, we lost the ability to detect missed interrupts. Since, we
have had issues with interrupts on a number of generations, and they are
required to be delivered in a timely fashion for a smooth UX, it is
important that we do log errors found in the wild and prevent the
display stalling for upwards of 1s every time the seqno interrupt is
missed.

Rather than continue to fix up the timeouts to work around the interface
impedence in wait_event_*(), open code the combination of
wait_event[_interruptible][_timeout], and use the exposed timer to
poll for seqno should we detect a lost interrupt.

v2: In order to satisfy the debug requirement of logging missed
interrupts with the real world requirments of making machines work even
if interrupts are hosed, we revert to polling after detecting a missed
interrupt.

v3: Throw in a debugfs interface to simulate broken hw not reporting
interrupts.

v4: s/EGAIN/EAGAIN/ (Imre)

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Imre Deak <imre.deak@intel.com>
[danvet: Don't use the struct typedef in new code.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

1 files changed, 67 insertions, 47 deletions

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 3ae925b0045..53e31513170 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -971,6 +971,17 @@ i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno)
 	return ret;
 }
 
+static void fake_irq(unsigned long data)
+{
+	wake_up_process((struct task_struct *)data);
+}
+
+static bool missed_irq(struct drm_i915_private *dev_priv,
+		       struct intel_ring_buffer *ring)
+{
+	return test_bit(ring->id, &dev_priv->gpu_error.missed_irq_rings);
+}
+
 /**
  * __wait_seqno - wait until execution of seqno has finished
  * @ring: the ring expected to report seqno
@@ -994,10 +1005,9 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			bool interruptible, struct timespec *timeout)
 {
 	drm_i915_private_t *dev_priv = ring->dev->dev_private;
-	struct timespec before, now, wait_time={1,0};
-	unsigned long timeout_jiffies;
-	long end;
-	bool wait_forever = true;
+	struct timespec before, now;
+	DEFINE_WAIT(wait);
+	long timeout_jiffies;
 	int ret;
 
 	WARN(dev_priv->pc8.irqs_disabled, "IRQs disabled\n");
@@ -1005,51 +1015,71 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
 		return 0;
 
-	trace_i915_gem_request_wait_begin(ring, seqno);
-
-	if (timeout != NULL) {
-		wait_time = *timeout;
-		wait_forever = false;
-	}
-
-	timeout_jiffies = timespec_to_jiffies_timeout(&wait_time);
+	timeout_jiffies = timeout ? timespec_to_jiffies_timeout(timeout) : 1;
 
-	if (WARN_ON(!ring->irq_get(ring)))
+	if (!(dev_priv->gpu_error.test_irq_rings & intel_ring_flag(ring)) &&
+	    WARN_ON(!ring->irq_get(ring)))
 		return -ENODEV;
 
-	/* Record current time in case interrupted by signal, or wedged * */
+	/* Record current time in case interrupted by signal, or wedged */
+	trace_i915_gem_request_wait_begin(ring, seqno);
 	getrawmonotonic(&before);
+	for (;;) {
+		struct timer_list timer;
+		unsigned long expire;
 
-#define EXIT_COND \
-	(i915_seqno_passed(ring->get_seqno(ring, false), seqno) || \
-	 i915_reset_in_progress(&dev_priv->gpu_error) || \
-	 reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
-	do {
-		if (interruptible)
-			end = wait_event_interruptible_timeout(ring->irq_queue,
-							       EXIT_COND,
-							       timeout_jiffies);
-		else
-			end = wait_event_timeout(ring->irq_queue, EXIT_COND,
-						 timeout_jiffies);
+		prepare_to_wait(&ring->irq_queue, &wait,
+				interruptible ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE);
 
 		/* We need to check whether any gpu reset happened in between
 		 * the caller grabbing the seqno and now ... */
-		if (reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
-			end = -EAGAIN;
+		if (reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter)) {
+			/* ... but upgrade the -EAGAIN to an -EIO if the gpu
+			 * is truely gone. */
+			ret = i915_gem_check_wedge(&dev_priv->gpu_error, interruptible);
+			if (ret == 0)
+				ret = -EAGAIN;
+			break;
+		}
 
-		/* ... but upgrade the -EGAIN to an -EIO if the gpu is truely
-		 * gone. */
-		ret = i915_gem_check_wedge(&dev_priv->gpu_error, interruptible);
-		if (ret)
-			end = ret;
-	} while (end == 0 && wait_forever);
+		if (i915_seqno_passed(ring->get_seqno(ring, false), seqno)) {
+			ret = 0;
+			break;
+		}
+
+		if (interruptible && signal_pending(current)) {
+			ret = -ERESTARTSYS;
+			break;
+		}
+
+		if (timeout_jiffies <= 0) {
+			ret = -ETIME;
+			break;
+		}
 
+		timer.function = NULL;
+		if (timeout || missed_irq(dev_priv, ring)) {
+			setup_timer_on_stack(&timer, fake_irq, (unsigned long)current);
+			expire = jiffies + (missed_irq(dev_priv, ring) ? 1: timeout_jiffies);
+			mod_timer(&timer, expire);
+		}
+
+		schedule();
+
+		if (timeout)
+			timeout_jiffies = expire - jiffies;
+
+		if (timer.function) {
+			del_singleshot_timer_sync(&timer);
+			destroy_timer_on_stack(&timer);
+		}
+	}
 	getrawmonotonic(&now);
+	trace_i915_gem_request_wait_end(ring, seqno);
 
 	ring->irq_put(ring);
-	trace_i915_gem_request_wait_end(ring, seqno);
-#undef EXIT_COND
+
+	finish_wait(&ring->irq_queue, &wait);
 
 	if (timeout) {
 		struct timespec sleep_time = timespec_sub(now, before);
@@ -1058,17 +1088,7 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			set_normalized_timespec(timeout, 0, 0);
 	}
 
-	switch (end) {
-	case -EIO:
-	case -EAGAIN: /* Wedged */
-	case -ERESTARTSYS: /* Signal */
-		return (int)end;
-	case 0: /* Timeout */
-		return -ETIME;
-	default: /* Completed */
-		WARN_ON(end < 0); /* We're not aware of other errors */
-		return 0;
-	}
+	return ret;
 }
 
 /**