dpdk-cryptodev: improve dequeue behavior, fix cache stats logging

This patch provides minor improvements to the logic governing dequeuing
from the ring. Previously whenever a frame was dequeued
we've been trying to dequeue from the ring another one till
inflight == 0. Now threshold is set for 8 frames pending in the cache
to be consumed by the vnet. This threshold has been chosen based on
cache ring stats observation in the system under load.
Some unnecessary logic for setting deq_tail has been removed.
Also logging has been corrected, and cache ring logic simplied.

Type: improvement
Signed-off-by: Piotr Bronowski <piotrx.bronowski@intel.com>
Change-Id: I19f3daf5913006e9cb23e142a163f596e85f5bda

1 files changed, 48 insertions, 25 deletions

diff --git a/src/plugins/dpdk/cryptodev/cryptodev.c b/src/plugins/dpdk/cryptodev/cryptodev.c
index 4e8bc026f58..43c2c879aab 100644
--- a/src/plugins/dpdk/cryptodev/cryptodev.c
+++ b/src/plugins/dpdk/cryptodev/cryptodev.c
@@ -672,50 +672,73 @@ cryptodev_show_cache_rings_fn (vlib_main_t *vm, unformat_input_t *input,
 {
   cryptodev_main_t *cmt = &cryptodev_main;
   u32 thread_index = 0;
+  u16 i;
   vec_foreach_index (thread_index, cmt->per_thread_data)
     {
       cryptodev_engine_thread_t *cet = cmt->per_thread_data + thread_index;
       cryptodev_cache_ring_t *ring = &cet->cache_ring;
       u16 head = ring->head;
       u16 tail = ring->tail;
-      u16 n_cached = ((head == tail) && (ring->frames[head].f == 0)) ?
-			     0 :
-		     ((head == tail) && (ring->frames[head].f != 0)) ?
-			     (CRYPTODEV_CACHE_QUEUE_MASK + 1) :
-		     (head > tail) ?
-			     (head - tail) :
-			     (CRYPTODEV_CACHE_QUEUE_MASK - tail + head);
+      u16 n_cached = (CRYPTODEV_CACHE_QUEUE_SIZE - tail + head) &
+		     CRYPTODEV_CACHE_QUEUE_MASK;
 
       u16 enq_head = ring->enq_head;
       u16 deq_tail = ring->deq_tail;
       u16 n_frames_inflight =
-	((enq_head == deq_tail) && (ring->frames[enq_head].f == 0)) ?
+	(enq_head == deq_tail) ?
 		0 :
-	((enq_head == deq_tail) && (ring->frames[enq_head].f != 0)) ?
-		CRYPTODEV_CACHE_QUEUE_MASK + 1 :
-	(enq_head > deq_tail) ?
-		(enq_head - deq_tail) :
-		(CRYPTODEV_CACHE_QUEUE_MASK - deq_tail + enq_head);
-
+		((CRYPTODEV_CACHE_QUEUE_SIZE + enq_head - deq_tail) &
+	   CRYPTODEV_CACHE_QUEUE_MASK);
+      /* even if some elements of dequeued frame are still pending for deq
+       * we consider the frame as processed */
       u16 n_frames_processed =
-	((tail == deq_tail) && (ring->frames[deq_tail].f == 0)) ? 0 :
-	((tail == deq_tail) && (ring->frames[deq_tail].f != 0)) ? 1 :
-	(deq_tail > tail) ? (deq_tail - tail + 1) :
-				  (CRYPTODEV_CACHE_QUEUE_MASK - tail + deq_tail - 1);
+	((tail == deq_tail) && (ring->frames[deq_tail].f == 0)) ?
+		0 :
+		((CRYPTODEV_CACHE_QUEUE_SIZE - tail + deq_tail) &
+	   CRYPTODEV_CACHE_QUEUE_MASK) +
+	    1;
+      /* even if some elements of enqueued frame are still pending for enq
+       * we consider the frame as enqueued */
+      u16 n_frames_pending =
+	(head == enq_head) ? 0 :
+				   ((CRYPTODEV_CACHE_QUEUE_SIZE - enq_head + head) &
+			      CRYPTODEV_CACHE_QUEUE_MASK) -
+			       1;
+
+      u16 elts_to_enq =
+	(ring->frames[enq_head].n_elts - ring->frames[enq_head].enq_elts_head);
+      u16 elts_to_deq =
+	(ring->frames[deq_tail].n_elts - ring->frames[deq_tail].deq_elts_tail);
+
+      u32 elts_total = 0;
+
+      for (i = 0; i < CRYPTODEV_CACHE_QUEUE_SIZE; i++)
+	elts_total += ring->frames[i].n_elts;
 
       if (vlib_num_workers () > 0 && thread_index == 0)
 	continue;
 
       vlib_cli_output (vm, "\n\n");
-      vlib_cli_output (vm, "Frames total: %u", n_cached);
-      vlib_cli_output (vm, "Frames pending in the ring: %u",
-		       n_cached - n_frames_inflight - n_frames_processed);
+      vlib_cli_output (vm, "Frames cached in the ring: %u", n_cached);
+      vlib_cli_output (vm, "Frames cached but not processed: %u",
+		       n_frames_pending);
       vlib_cli_output (vm, "Frames inflight: %u", n_frames_inflight);
-      vlib_cli_output (vm, "Frames dequed but not returned: %u",
-		       n_frames_processed);
+      vlib_cli_output (vm, "Frames processed: %u", n_frames_processed);
+      vlib_cli_output (vm, "Elements total: %u", elts_total);
       vlib_cli_output (vm, "Elements inflight: %u", cet->inflight);
-      vlib_cli_output (vm, "Head: %u", head);
-      vlib_cli_output (vm, "Tail: %u", tail);
+      vlib_cli_output (vm, "Head index: %u", head);
+      vlib_cli_output (vm, "Tail index: %u", tail);
+      vlib_cli_output (vm, "Current frame index beeing enqueued: %u",
+		       enq_head);
+      vlib_cli_output (vm, "Current frame index being dequeued: %u", deq_tail);
+      vlib_cli_output (vm,
+		       "Elements in current frame to be enqueued: %u, waiting "
+		       "to be enqueued: %u",
+		       ring->frames[enq_head].n_elts, elts_to_enq);
+      vlib_cli_output (vm,
+		       "Elements in current frame to be dequeued: %u, waiting "
+		       "to be dequeued: %u",
+		       ring->frames[deq_tail].n_elts, elts_to_deq);
       vlib_cli_output (vm, "\n\n");
     }
   return 0;