[dpdk-dev] net/tap: fix the vdev data sharing for tun
Checks
Commit Message
Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
Enables TUN PMD sharing by attaching the port from the shared data.
Cc: zhihong.wang@intel.com
Cc: ferruh.yigit@intel.com
Signed-off-by: Vipin Varghese <vipin.varghese@intel.com>
---
---
drivers/net/tap/rte_eth_tap.c | 13 +++++++++++++
1 file changed, 13 insertions(+)
Comments
> On May 12, 2018, at 1:37 AM, Vipin Varghese <vipin.varghese@intel.com> wrote:
>
> Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
>
> Enables TUN PMD sharing by attaching the port from the shared data.
>
> Cc: zhihong.wang@intel.com
> Cc: ferruh.yigit@intel.com
>
> Signed-off-by: Vipin Varghese <vipin.varghese@intel.com>
> ---
> ---
> drivers/net/tap/rte_eth_tap.c | 13 +++++++++++++
> 1 file changed, 13 insertions(+)
>
> diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c
> index db397d6..ea6d899 100644
> --- a/drivers/net/tap/rte_eth_tap.c
> +++ b/drivers/net/tap/rte_eth_tap.c
> @@ -1657,6 +1657,7 @@ rte_pmd_tun_probe(struct rte_vdev_device *dev)
> struct rte_kvargs *kvlist = NULL;
> char tun_name[RTE_ETH_NAME_MAX_LEN];
> char remote_iface[RTE_ETH_NAME_MAX_LEN];
> + struct rte_eth_dev *eth_dev;
>
> tap_type = 0;
> strcpy(tuntap_name, "TUN");
> @@ -1665,6 +1666,18 @@ rte_pmd_tun_probe(struct rte_vdev_device *dev)
> params = rte_vdev_device_args(dev);
> memset(remote_iface, 0, RTE_ETH_NAME_MAX_LEN);
>
> + if (rte_eal_process_type() == RTE_PROC_SECONDARY &&
> + strlen(params) == 0) {
> + eth_dev = rte_eth_dev_attach_secondary(name);
> + if (!eth_dev) {
> + TAP_LOG(ERR, "Failed to probe %s", name);
> + return -1;
> + }
> + /* TODO: request info from primary to set up Rx and Tx */
What is the rule (if we have one) about having TODO, FIXME, … like comments in the code? Maybe remove the todo and create a enhancement request for later.
> + eth_dev->dev_ops = &ops;
> + return 0;
> + }
> +
> snprintf(tun_name, sizeof(tun_name), "%s%u",
> DEFAULT_TUN_NAME, tun_unit++);
>
> --
> 2.7.4
>
Regards,
Keith
HI Keith,
Thanks for the input, please find my answer inline to email therad
<Snipped>
> > + /* TODO: request info from primary to set up Rx and Tx */
>
> What is the rule (if we have one) about having TODO, FIXME, … like comments in
> the code? Maybe remove the todo and create a enhancement request for later.
>
I am not aware of the 'To Do' planned. This came in as part of logic block from previous commit for all PMD to support Secondary
<Snipped>
>
> Regards,
> Keith
On 5/15/2018 1:56 PM, Varghese, Vipin wrote:
> HI Keith,
>
> Thanks for the input, please find my answer inline to email therad
>
> <Snipped>
>
>>> + /* TODO: request info from primary to set up Rx and Tx */
>>
>> What is the rule (if we have one) about having TODO, FIXME, … like comments in
>> the code? Maybe remove the todo and create a enhancement request for later.
>>
>
> I am not aware of the 'To Do' planned. This came in as part of logic block from previous commit for all PMD to support Secondary
Specially when there is a work not planned for this, agreed to drop todo, can
you please send a new version without that note please?
>
> <Snipped>
>
>>
>> Regards,
>> Keith
>
@@ -1657,6 +1657,7 @@ rte_pmd_tun_probe(struct rte_vdev_device *dev)
struct rte_kvargs *kvlist = NULL;
char tun_name[RTE_ETH_NAME_MAX_LEN];
char remote_iface[RTE_ETH_NAME_MAX_LEN];
+ struct rte_eth_dev *eth_dev;
tap_type = 0;
strcpy(tuntap_name, "TUN");
@@ -1665,6 +1666,18 @@ rte_pmd_tun_probe(struct rte_vdev_device *dev)
params = rte_vdev_device_args(dev);
memset(remote_iface, 0, RTE_ETH_NAME_MAX_LEN);
+ if (rte_eal_process_type() == RTE_PROC_SECONDARY &&
+ strlen(params) == 0) {
+ eth_dev = rte_eth_dev_attach_secondary(name);
+ if (!eth_dev) {
+ TAP_LOG(ERR, "Failed to probe %s", name);
+ return -1;
+ }
+ /* TODO: request info from primary to set up Rx and Tx */
+ eth_dev->dev_ops = &ops;
+ return 0;
+ }
+
snprintf(tun_name, sizeof(tun_name), "%s%u",
DEFAULT_TUN_NAME, tun_unit++);