drm/i915: Avoid keeping waitboost active for signaling threads

Once a client has requested a waitboost, we keep that waitboost active
until all clients are no longer waiting. This is because we don't
distinguish which waiter deserves the boost. However, with the advent of
fence signaling, the signaler threads appear as waiters to the RPS
interrupt handler. So instead of using a single boolean to track when to
keep the waitboost active, use a counter of all outstanding waitboosted
requests.

At this point, I have removed all vestiges of the rate limiting on
clients. Whilst this means that compositors should remain more fluid,
it also means that boosts are more prevalent. See commit b29c19b64528
("drm/i915: Boost RPS frequency for CPU stalls") for a longer discussion
on the pros and cons of both approaches.

A drawback of this implementation is that it requires constant request
submission to keep the waitboost trimmed (as it is now cancelled when the
request is completed). This will be fine for a busy system, but near
idle the boosts may be kept for longer than desired (effectively tens of
vblanks worstcase) and there is a reliance on rc6 instead.

v2: Remove defunct rps.client_lock

Reported-by: Michał Winiarski <michal.winiarski@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michał Winiarski <michal.winiarski@intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170628123548.9236-1-chris@chris-wilson.co.uk

1 files changed, 3 insertions, 15 deletions

diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index e4934d5..1d33cea 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -1091,18 +1091,6 @@ static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir)
 	return events;
 }
 
-static bool any_waiters(struct drm_i915_private *dev_priv)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-
-	for_each_engine(engine, dev_priv, id)
-		if (intel_engine_has_waiter(engine))
-			return true;
-
-	return false;
-}
-
 static void gen6_pm_rps_work(struct work_struct *work)
 {
 	struct drm_i915_private *dev_priv =
@@ -1114,7 +1102,7 @@ static void gen6_pm_rps_work(struct work_struct *work)
 	spin_lock_irq(&dev_priv->irq_lock);
 	if (dev_priv->rps.interrupts_enabled) {
 		pm_iir = fetch_and_zero(&dev_priv->rps.pm_iir);
-		client_boost = fetch_and_zero(&dev_priv->rps.client_boost);
+		client_boost = atomic_read(&dev_priv->rps.num_waiters);
 	}
 	spin_unlock_irq(&dev_priv->irq_lock);
 
@@ -1131,7 +1119,7 @@ static void gen6_pm_rps_work(struct work_struct *work)
 	new_delay = dev_priv->rps.cur_freq;
 	min = dev_priv->rps.min_freq_softlimit;
 	max = dev_priv->rps.max_freq_softlimit;
-	if (client_boost || any_waiters(dev_priv))
+	if (client_boost)
 		max = dev_priv->rps.max_freq;
 	if (client_boost && new_delay < dev_priv->rps.boost_freq) {
 		new_delay = dev_priv->rps.boost_freq;
@@ -1144,7 +1132,7 @@ static void gen6_pm_rps_work(struct work_struct *work)
 
 		if (new_delay >= dev_priv->rps.max_freq_softlimit)
 			adj = 0;
-	} else if (client_boost || any_waiters(dev_priv)) {
+	} else if (client_boost) {
 		adj = 0;
 	} else if (pm_iir & GEN6_PM_RP_DOWN_TIMEOUT) {
 		if (dev_priv->rps.cur_freq > dev_priv->rps.efficient_freq)