-----Original Message-----
From: Van Haaren, Harry
Sent: Friday, March 23, 2018 12:06 PM
To: Eads, Gage <redacted>; dev@dpdk.org
Cc: jerin.jacob@caviumnetworks.com; hemant.agrawal@nxp.com; Richardson,
Bruce [off-list ref]; santosh.shukla@caviumnetworks.com;
nipun.gupta@nxp.com
Subject: RE: [PATCH v4 2/2] event/sw: support device stop flush callback

quoted

From: Eads, Gage
Sent: Tuesday, March 20, 2018 2:13 PM
To: dev@dpdk.org
Cc: jerin.jacob@caviumnetworks.com; Van Haaren, Harry
[off-list ref]; hemant.agrawal@nxp.com; Richardson,
Bruce [off-list ref]; santosh.shukla@caviumnetworks.com;
nipun.gupta@nxp.com
Subject: [PATCH v4 2/2] event/sw: support device stop flush callback

This commit also adds a flush callback test to the sw eventdev's
selftest suite.

Signed-off-by: Gage Eads <redacted>
---
 drivers/event/sw/sw_evdev.c          | 25 ++++++++++-
 drivers/event/sw/sw_evdev_selftest.c | 80
+++++++++++++++++++++++++++++++++++-
 2 files changed, 102 insertions(+), 3 deletions(-)

diff --git a/drivers/event/sw/sw_evdev.c b/drivers/event/sw/sw_evdev.c
index 0e89f11..11f394f 100644
--- a/drivers/event/sw/sw_evdev.c
+++ b/drivers/event/sw/sw_evdev.c

@@ -362,8 +362,25 @@ sw_init_qid_iqs(struct sw_evdev *sw)  }

 static void
-sw_clean_qid_iqs(struct sw_evdev *sw)
+sw_flush_iq(struct rte_eventdev *dev, struct sw_iq *iq)
 {
+	struct sw_evdev *sw = sw_pmd_priv(dev);
+
+	while (iq_count(iq) > 0) {
+		struct rte_event event;
+
+		iq_dequeue_burst(sw, iq, &event, 1);
+
+		dev->dev_ops->dev_stop_flush(dev->data->dev_id,
+					     event,
+					     dev->data->dev_stop_flush_arg);

Adding check that dev_stop_flush() is non-NULL?

[Update: Ah I see you do this below already. Still, better check twice I think, the
data path isn't running here anyway in case future me decides to call
sw_flush_iq() without performing the check]

Agreed, will fix in v5.

if(dev->dev_ops->dev_stop_flush)
   dev->dev_ops->dev_stop_flush(...);

quoted

@@ -702,7 +723,7 @@ static void
 sw_stop(struct rte_eventdev *dev)
 {
 	struct sw_evdev *sw = sw_pmd_priv(dev);
-	sw_clean_qid_iqs(sw);
+	sw_clean_qid_iqs(dev);

Based on the port buffers comment on 1/2, we probably need a
sw_clean_port_buffers(sw); here to return any events in the port owned SW
buffers?

Agreed, if there is buy-in on the approach discussed in patch #1.