[Intel-gfx] [PATCH v4 07/15] drm/i915/perf: better pipeline aged/aging tail updates

[Robert Bragg]

This updates the tail pointer race workaround handling to updating the
'aged' pointer before looking to start aging a new one. There's the
possibility that there is already new data available and so we can
immediately start aging a new pointer without having to first wait for a
later hrtimer callback (and then another to age).

Signed-off-by: Robert Bragg <robert at sixbynine.org>
Reviewed-by: Matthew Auld <matthew.auld at intel.com>
---
 drivers/gpu/drm/i915/i915_perf.c | 41 ++++++++++++++++++++++------------------
 1 file changed, 23 insertions(+), 18 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 08cc2b0dd734..5738b99caa5b 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -391,6 +391,29 @@ static bool gen7_oa_buffer_check_unlocked(struct drm_i915_private *dev_priv)
 
 	now = ktime_get_mono_fast_ns();
 
+	/* Update the aged tail
+	 *
+	 * Flip the tail pointer available for read()s once the aging tail is
+	 * old enough to trust that the corresponding data will be visible to
+	 * the CPU...
+	 *
+	 * Do this before updating the aging pointer in case we may be able to
+	 * immediately start aging a new pointer too (if new data has become
+	 * available) without needing to wait for a later hrtimer callback.
+	 */
+	if (aging_tail != INVALID_TAIL_PTR &&
+	    ((now - dev_priv->perf.oa.oa_buffer.aging_timestamp) >
+	     OA_TAIL_MARGIN_NSEC)) {
+		aged_idx ^= 1;
+		dev_priv->perf.oa.oa_buffer.aged_tail_idx = aged_idx;
+
+		aged_tail = aging_tail;
+
+		/* Mark that we need a new pointer to start aging... */
+		dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = INVALID_TAIL_PTR;
+		aging_tail = INVALID_TAIL_PTR;
+	}
+
 	/* Update the aging tail
 	 *
 	 * We throttle aging tail updates until we have a new tail that
@@ -420,24 +443,6 @@ static bool gen7_oa_buffer_check_unlocked(struct drm_i915_private *dev_priv)
 		}
 	}
 
-	/* Update the aged tail
-	 *
-	 * Flip the tail pointer available for read()s once the aging tail is
-	 * old enough to trust that the corresponding data will be visible to
-	 * the CPU...
-	 */
-	if (aging_tail != INVALID_TAIL_PTR &&
-	    ((now - dev_priv->perf.oa.oa_buffer.aging_timestamp) >
-	     OA_TAIL_MARGIN_NSEC)) {
-		aged_idx ^= 1;
-		dev_priv->perf.oa.oa_buffer.aged_tail_idx = aged_idx;
-
-		aged_tail = aging_tail;
-
-		/* Mark that we need a new pointer to start aging... */
-		dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = INVALID_TAIL_PTR;
-	}
-
 	spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
 
 	return aged_tail == INVALID_TAIL_PTR ?
-- 
2.12.0