[dpdk-dev] [PATCH] net/virtio-user: fix add pointer checking for rxvq

[dpdk-dev] [PATCH] net/virtio-user: fix add pointer checking for rxvq

[zhiyong.yang]

For virtio-user server mode, One use case comes across segmentation fault.
step 1: Launch vhost side as client firstly.
step 2: launch virtio-user side as server.

The cause is: after registering virtio_interrupt_handler into
eal-intr-thread, two threads (main thread and eal-intr-thread) have
sync issues, so add rxvq pointer checking in function virtio_notify_peers
to decide if the code can continue.

Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
Cc: stable@dpdk.org
Cc: maxime.coquelin@redhat.com
Cc: tiwei.bie@intel.com
Cc: lei.a.yao@intel.com
Cc: ferruh.yigit@intel.com

Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
---
 drivers/net/virtio/virtio_ethdev.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c
index 5833dad73..df50a571a 100644
--- a/drivers/net/virtio/virtio_ethdev.c
+++ b/drivers/net/virtio/virtio_ethdev.c
@@ -1245,6 +1245,9 @@ virtio_notify_peers(struct rte_eth_dev *dev)
 		return;
 
 	rxvq = dev->data->rx_queues[0];
+	if (!rxvq)
+		return;
+
 	rarp_mbuf = rte_net_make_rarp_packet(rxvq->mpool,
 			(struct ether_addr *)hw->mac_addr);
 	if (rarp_mbuf == NULL) {
-- 
2.14.3

Re: [dpdk-dev] [PATCH] net/virtio-user: fix add pointer checking for rxvq

[Maxime Coquelin]

Hi,

I would use net/virtio prefix instead, as IIUC, the impacted code can
also be called by Virtio PMD.

On 05/24/2018 03:04 PM, zhiyong.yang@intel.com wrote:
> For virtio-user server mode, One use case comes across segmentation fault.
> step 1: Launch vhost side as client firstly.
> step 2: launch virtio-user side as server.
> 
> The cause is: after registering virtio_interrupt_handler into
> eal-intr-thread, two threads (main thread and eal-intr-thread) have
> sync issues, so add rxvq pointer checking in function virtio_notify_peers
> to decide if the code can continue.
> 
> Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
> Cc: stable@dpdk.org
> Cc: maxime.coquelin@redhat.com
> Cc: tiwei.bie@intel.com
> Cc: lei.a.yao@intel.com
> Cc: ferruh.yigit@intel.com
> 
> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
> ---
>   drivers/net/virtio/virtio_ethdev.c | 3 +++
>   1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c
> index 5833dad73..df50a571a 100644
> --- a/drivers/net/virtio/virtio_ethdev.c
> +++ b/drivers/net/virtio/virtio_ethdev.c
> @@ -1245,6 +1245,9 @@ virtio_notify_peers(struct rte_eth_dev *dev)
>   		return;
>   
>   	rxvq = dev->data->rx_queues[0];
> +	if (!rxvq)
> +		return;
> +

Other than that, it looks good to me:
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>

Thanks,
Maxime

>   	rarp_mbuf = rte_net_make_rarp_packet(rxvq->mpool,
>   			(struct ether_addr *)hw->mac_addr);
>   	if (rarp_mbuf == NULL) {
> 

[dpdk-dev] [PATCH v2] net/virtio: fix add pointer checking for rxvq

[zhiyong.yang]

For virtio-user server mode, one use case comes across segmentation fault.
step 1: Launch vhost side as client firstly.
step 2: launch virtio-user side as server.

The cause is: after registering virtio_interrupt_handler into
eal-intr-thread, two threads (main thread and eal-intr-thread) have
sync issues, so add rxvq pointer checking in function virtio_notify_peers
to decide if the code can continue.

Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
Cc: stable@dpdk.org
Cc: maxime.coquelin@redhat.com
Cc: tiwei.bie@intel.com
Cc: lei.a.yao@intel.com
Cc: ferruh.yigit@intel.com

Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
---

Change in V2:
change title prefix from net/virtio-user to net/virtio.

 drivers/net/virtio/virtio_ethdev.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c
index 5833dad73..df50a571a 100644
--- a/drivers/net/virtio/virtio_ethdev.c
+++ b/drivers/net/virtio/virtio_ethdev.c
@@ -1245,6 +1245,9 @@ virtio_notify_peers(struct rte_eth_dev *dev)
 		return;
 
 	rxvq = dev->data->rx_queues[0];
+	if (!rxvq)
+		return;
+
 	rarp_mbuf = rte_net_make_rarp_packet(rxvq->mpool,
 			(struct ether_addr *)hw->mac_addr);
 	if (rarp_mbuf == NULL) {
-- 
2.14.3

Re: [dpdk-dev] [PATCH v2] net/virtio: fix add pointer checking for rxvq

[Maxime Coquelin]

On 05/24/2018 05:51 PM, zhiyong.yang@intel.com wrote:
> For virtio-user server mode, one use case comes across segmentation fault.
> step 1: Launch vhost side as client firstly.
> step 2: launch virtio-user side as server.
> 
> The cause is: after registering virtio_interrupt_handler into
> eal-intr-thread, two threads (main thread and eal-intr-thread) have
> sync issues, so add rxvq pointer checking in function virtio_notify_peers
> to decide if the code can continue.
> 
> Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
> Cc: stable@dpdk.org
> Cc: maxime.coquelin@redhat.com
> Cc: tiwei.bie@intel.com
> Cc: lei.a.yao@intel.com
> Cc: ferruh.yigit@intel.com
> 
> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
> ---
> 
> Change in V2:
> change title prefix from net/virtio-user to net/virtio.
> 
>   drivers/net/virtio/virtio_ethdev.c | 3 +++
>   1 file changed, 3 insertions(+)

Applied to dpdk-next-virtio/master.

Thanks,
Maxime