[dpdk-stable] [PATCH v1 1/2] app/eventdev: adjust event count order for pipeline test

[dpdk-stable] [PATCH v1 1/2] app/eventdev: adjust event count order for pipeline test

[Feifei Wang]

For the fwd mode (internal_port = false) in pipeline test,
processed-pkts increment should after enqueue. However, in
multi_stage_fwd and multi_stage_burst_fwd, "w->processed_pkts" is
increased before enqueue.

To fix this, move "w->processed_pkts" increment after enqueue, and then
the main core can load the correct number of processed packets.

Fixes: 314bcf58ca8f ("app/eventdev: add pipeline queue worker functions")
Cc: pbhagavatula@marvell.com
Cc: stable@dpdk.org

Signed-off-by: Feifei Wang <feifei.wang2@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
---
 app/test-eventdev/test_pipeline_queue.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/app/test-eventdev/test_pipeline_queue.c b/app/test-eventdev/test_pipeline_queue.c
index 7bebac34f..01f33e3b4 100644
--- a/app/test-eventdev/test_pipeline_queue.c
+++ b/app/test-eventdev/test_pipeline_queue.c
@@ -180,13 +180,13 @@ pipeline_queue_worker_multi_stage_fwd(void *arg)
 			ev.queue_id = tx_queue[ev.mbuf->port];
 			rte_event_eth_tx_adapter_txq_set(ev.mbuf, 0);
 			pipeline_fwd_event(&ev, RTE_SCHED_TYPE_ATOMIC);
+			pipeline_event_enqueue(dev, port, &ev);
 			w->processed_pkts++;
 		} else {
 			ev.queue_id++;
 			pipeline_fwd_event(&ev, sched_type_list[cq_id]);
+			pipeline_event_enqueue(dev, port, &ev);
 		}
-
-		pipeline_event_enqueue(dev, port, &ev);
 	}
 
 	return 0;
@@ -237,6 +237,7 @@ pipeline_queue_worker_multi_stage_burst_fwd(void *arg)
 	const uint8_t *tx_queue = t->tx_evqueue_id;
 
 	while (t->done == false) {
+		uint16_t processed_pkts = 0;
 		uint16_t nb_rx = rte_event_dequeue_burst(dev, port, ev,
 				BURST_SIZE, 0);
 
@@ -254,7 +255,7 @@ pipeline_queue_worker_multi_stage_burst_fwd(void *arg)
 				rte_event_eth_tx_adapter_txq_set(ev[i].mbuf, 0);
 				pipeline_fwd_event(&ev[i],
 						RTE_SCHED_TYPE_ATOMIC);
-				w->processed_pkts++;
+				processed_pkts++;
 			} else {
 				ev[i].queue_id++;
 				pipeline_fwd_event(&ev[i],
@@ -263,6 +264,7 @@ pipeline_queue_worker_multi_stage_burst_fwd(void *arg)
 		}
 
 		pipeline_event_enqueue_burst(dev, port, ev, nb_rx);
+		w->processed_pkts += processed_pkts;
 	}
 
 	return 0;
-- 
2.25.1

[dpdk-stable] [PATCH v1 2/2] app/eventdev: remove redundant enqueue in burst Tx

[Feifei Wang]

For eventdev pipeline test, in burst_tx cases, there is no needed to
set ev.op as RTE_EVENT_OP_RELEASE and call pipeline_event_enqueue_burst
to release events. This is because for tx mode(internal_port=true),
the capability "implicit_release" of dev is enabled, and the app can
release events by "rte_event_dequeue_burst" rather than enqueue.

Fixes: 314bcf58ca8f ("app/eventdev: add pipeline queue worker functions")
Cc: pbhagavatula@marvell.com
Cc: stable@dpdk.org

Signed-off-by: Feifei Wang <feifei.wang2@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
---
 app/test-eventdev/test_pipeline_queue.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/app/test-eventdev/test_pipeline_queue.c b/app/test-eventdev/test_pipeline_queue.c
index 01f33e3b4..9a9febb19 100644
--- a/app/test-eventdev/test_pipeline_queue.c
+++ b/app/test-eventdev/test_pipeline_queue.c
@@ -83,16 +83,15 @@ pipeline_queue_worker_single_stage_burst_tx(void *arg)
 			rte_prefetch0(ev[i + 1].mbuf);
 			if (ev[i].sched_type == RTE_SCHED_TYPE_ATOMIC) {
 				pipeline_event_tx(dev, port, &ev[i]);
-				ev[i].op = RTE_EVENT_OP_RELEASE;
 				w->processed_pkts++;
 			} else {
 				ev[i].queue_id++;
 				pipeline_fwd_event(&ev[i],
 						RTE_SCHED_TYPE_ATOMIC);
+				pipeline_event_enqueue_burst(dev, port, ev,
+						nb_rx);
 			}
 		}
-
-		pipeline_event_enqueue_burst(dev, port, ev, nb_rx);
 	}
 
 	return 0;
@@ -213,7 +212,6 @@ pipeline_queue_worker_multi_stage_burst_tx(void *arg)
 
 			if (ev[i].queue_id == tx_queue[ev[i].mbuf->port]) {
 				pipeline_event_tx(dev, port, &ev[i]);
-				ev[i].op = RTE_EVENT_OP_RELEASE;
 				w->processed_pkts++;
 				continue;
 			}
@@ -222,9 +220,8 @@ pipeline_queue_worker_multi_stage_burst_tx(void *arg)
 			pipeline_fwd_event(&ev[i], cq_id != last_queue ?
 					sched_type_list[cq_id] :
 					RTE_SCHED_TYPE_ATOMIC);
+			pipeline_event_enqueue_burst(dev, port, ev, nb_rx);
 		}
-
-		pipeline_event_enqueue_burst(dev, port, ev, nb_rx);
 	}
 
 	return 0;
-- 
2.25.1