Skip to contentSkip to conversation

ethdev: fix checking Rx/Tx queue status

PATCH Applied Yanglong Wu 4 participants 8 messages May 11, 2018

lore
Share & tools
lore

Yanglong Wu

author
#0 lore raw

PATCH 1/1 ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
lib/librte_ethdev/rte_ethdev.c +6 -4
index a357ee09f..34379642a 100644
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	txq = dev->data->tx_queues;
Signature · 1 line
2.11.0
#1 lore raw
Quoted 37 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
lib/librte_ethdev/rte_ethdev.c +3 -2
index a357ee09f..34379642a 100644
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id, uint16_t rx_queue_id,
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -3
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.
Why not fix offending PMDs instead?

Quoted 1 line

2.11.0

#2 lore raw
Quoted 58 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

lib/librte_ethdev/rte_ethdev.c rte_eth_rx_queue_setup(uint16_t port_id,

Quoted 22 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/lib/librte_ethdev/rte_ethdev.c index a357ee09f..34379642a 100644

lib/librte_ethdev/rte_ethdev.c +0 -0
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id,

uint16_t rx_queue_id,

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id,

uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
Quoted 4 lines

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +0 -1
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.

Em... would you explain more, I can't figure out why this break runtime queue setup so far.

Quoted 1 line

Why not fix offending PMDs instead?

Ixgbe, i40e have this issue, not sure what about others. probably we should fix at rte_eth_dev_stop to save all PMD?

Quoted 2 lines
Quoted 2 lines

2.11.0

Acked-by
#3 lore raw
Quoted 67 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 10:24 AM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 56 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

lib/librte_ethdev/rte_ethdev.c rte_eth_rx_queue_setup(uint16_t port_id,

Quoted 22 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/lib/librte_ethdev/rte_ethdev.c index a357ee09f..34379642a 100644

lib/librte_ethdev/rte_ethdev.c +0 -0
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id,

uint16_t rx_queue_id,

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id,

uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
Quoted 6 lines
Quoted 4 lines

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +0 -1
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.

Em... would you explain more, I can't figure out why this break runtime queue setup so far.

Ah yes, it was me not reading code properly.
So yep - you right it wouldn't break things.

Quoted 5 lines
Quoted 2 lines

Why not fix offending PMDs instead?

Ixgbe, i40e have this issue, not sure what about others. probably we should fix at rte_eth_dev_stop to save all PMD?

Probably yes, that's a good thing to do in long term anyway.

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

2.11.0

Acked-by Konstantin Ananyev <[email protected]>
#4 lore raw
Quoted 95 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 5:51 PM
To: Zhang, Qi Z <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 70 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 10:24 AM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 58 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

lib/librte_ethdev/rte_ethdev.c rte_eth_rx_queue_setup(uint16_t port_id,

Quoted 23 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/lib/librte_ethdev/rte_ethdev.c index a357ee09f..34379642a 100644

lib/librte_ethdev/rte_ethdev.c +0 -0
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id,

uint16_t rx_queue_id,

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id,

uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
Quoted 6 lines
Quoted 4 lines

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +0 -1
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.

Em... would you explain more, I can't figure out why this break runtime

queue setup so far.

Ah yes, it was me not reading code properly.
So yep - you right it wouldn't break things.

Quoted 1 line
Quoted 6 lines
Quoted 4 lines
Quoted 1 line

Why not fix offending PMDs instead?

Ixgbe, i40e have this issue, not sure what about others. probably we should fix at rte_eth_dev_stop to save all PMD?

Probably yes, that's a good thing to do in long term anyway.

I'm thinking if it's better to move all queue state change into ethdev layer.
So there could be a patch to update queue states properly at dev_start/dev_stop/queue_start/queue_stop and maintainer help them self to cleanup PMD, and I think our proposed fix should be part of this. But for now, can we just keep current simple fix?

Regards
Qi

Quoted 6 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

2.11.0

Acked-by Konstantin Ananyev <[email protected]>
#5 lore raw
Quoted 109 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 1:34 PM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 98 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 5:51 PM
To: Zhang, Qi Z <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 70 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 10:24 AM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 58 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

lib/librte_ethdev/rte_ethdev.c rte_eth_rx_queue_setup(uint16_t port_id,

Quoted 23 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/lib/librte_ethdev/rte_ethdev.c index a357ee09f..34379642a 100644

lib/librte_ethdev/rte_ethdev.c +0 -0
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id,

uint16_t rx_queue_id,

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id,

uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
Quoted 6 lines
Quoted 4 lines

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +0 -1
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.

Em... would you explain more, I can't figure out why this break runtime

queue setup so far.

Ah yes, it was me not reading code properly.
So yep - you right it wouldn't break things.

Quoted 5 lines
Quoted 2 lines

Why not fix offending PMDs instead?

Ixgbe, i40e have this issue, not sure what about others. probably we should fix at rte_eth_dev_stop to save all PMD?

Quoted 6 lines
Quoted 1 line

Probably yes, that's a good thing to do in long term anyway.

I'm thinking if it's better to move all queue state change into ethdev layer.
So there could be a patch to update queue states properly at dev_start/dev_stop/queue_start/queue_stop and maintainer help them self to cleanup PMD, and I think our proposed fix should be part of this.
But for now, can we just keep current simple fix?

Yes, I think that's a good one, that's why I acked it, see below. Konstantin

Quoted 10 lines

Regards
Qi

Quoted 6 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

2.11.0

Acked-by Konstantin Ananyev <[email protected]>
#6 lore raw
Quoted 129 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 8:53 PM
To: Zhang, Qi Z <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 116 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 1:34 PM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 103 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 5:51 PM
To: Zhang, Qi Z <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 72 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 10:24 AM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 60 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

lib/librte_ethdev/rte_ethdev.c rte_eth_rx_queue_setup(uint16_t port_id,

Quoted 25 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
---
 lib/librte_ethdev/rte_ethdev.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
(quoted diff) +0 -0

b/lib/librte_ethdev/rte_ethdev.c index a357ee09f..34379642a 100644

lib/librte_ethdev/rte_ethdev.c +0 -0
@@ -1479,8 +1479,9 @@ rte_eth_rx_queue_setup(uint16_t port_id,

uint16_t rx_queue_id,

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 12 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->rx_queue_state[rx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->rx_queue_state[rx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
 
 	rxq = dev->data->rx_queues;

@@ -1611,8 +1612,9 @@ rte_eth_tx_queue_setup(uint16_t port_id,

uint16_t tx_queue_id,

Quoted 10 lines
lib/librte_ethdev/rte_ethdev.c +3 -2
 			RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP))
 		return -EBUSY;
 
-	if (dev->data->tx_queue_state[tx_queue_id] !=
-		RTE_ETH_QUEUE_STATE_STOPPED)
+	if (dev->data->dev_started &&
+		(dev->data->tx_queue_state[tx_queue_id] !=
+			RTE_ETH_QUEUE_STATE_STOPPED))
 		return -EBUSY;
Quoted 7 lines
Quoted 4 lines

lib/librte_ethdev/rte_ethdev.c rte_eth_tx_queue_setup(uint16_t port_id,

Quoted 2 lines
lib/librte_ethdev/rte_ethdev.c +0 -1
 	txq = dev->data->tx_queues;
--

That would break runtime queue setup.

Em... would you explain more, I can't figure out why this break runtime

queue setup so far.

Ah yes, it was me not reading code properly.
So yep - you right it wouldn't break things.

Quoted 5 lines
Quoted 2 lines

Why not fix offending PMDs instead?

Ixgbe, i40e have this issue, not sure what about others. probably we should fix at rte_eth_dev_stop to save all PMD?

Probably yes, that's a good thing to do in long term anyway.

Quoted 8 lines
Quoted 3 lines

I'm thinking if it's better to move all queue state change into ethdev layer. So there could be a patch to update queue states properly at dev_start/dev_stop/queue_start/queue_stop and maintainer help them self

to cleanup PMD, and I think our proposed fix should be part of this.

Quoted 1 line

But for now, can we just keep current simple fix?

Yes, I think that's a good one, that's why I acked it, see below. Konstantin

oh, I missed that :)

Quoted 11 lines
Quoted 11 lines

Regards
Qi

Quoted 6 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

2.11.0

Acked-by Konstantin Ananyev <[email protected]>
Applied
#7 lore raw
In reply to Ananyev, Konstantin
Quoted 29 lines · Konstantin
Quoted 29 lines

-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, May 11, 2018 10:24 AM
To: Ananyev, Konstantin <[email protected]>; Wu, Yanglong <[email protected]>; [email protected]
Cc: Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 18 lines

-----Original Message-----
From: Ananyev, Konstantin
Sent: Friday, May 11, 2018 4:39 PM
To: Wu, Yanglong <[email protected]>; [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: RE: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 8 lines

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Yanglong Wu
Sent: Friday, May 11, 2018 9:22 AM
To: [email protected]
Cc: Zhang, Qi Z <[email protected]>; Lu, Wenzhuo <[email protected]>; Wu, Yanglong <[email protected]>
Subject: [dpdk-dev] [PATCH] ethdev: fix checking Rx/Tx queue status

Quoted 7 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

Relax the check for queue setup, since some device may not update queue states during dev_stop.

Fixes cac923cfea47 ("ethdev: support runtime queue setup")
Signed-off-by Yanglong Wu <[email protected]>
Acked-by Konstantin Ananyev <[email protected]>

Applied to dpdk-next-net/master, thanks.