[dpdk-dev] [PATCH v2 04/13] ethdev: change promiscuous callbacks to return status

Hyong Youb Kim (hyonkim) hyonkim at cisco.com
Mon Sep 16 08:48:42 CEST 2019


> -----Original Message-----
> From: Andrew Rybchenko <arybchenko at solarflare.com>
> Sent: Saturday, September 14, 2019 6:07 AM
[...]
> Subject: Re: [PATCH v2 04/13] ethdev: change promiscuous callbacks to
> return status
> 
> On 9/11/19 11:46 AM, Hyong Youb Kim (hyonkim) wrote:
> >> -----Original Message-----
> >> From: Andrew Rybchenko <arybchenko at solarflare.com>
> >> Sent: Monday, September 9, 2019 8:59 PM
> > [...]
> >> Subject: [PATCH v2 04/13] ethdev: change promiscuous callbacks to return
> >> status
> >>
> >> Enabling/disabling of promiscuous mode is not always successful and
> >> it should be taken into account to be able to handle it properly.
> >>
> >> When correct return status is unclear from driver code, -EAGAIN is used.
> >>
> >> Signed-off-by: Andrew Rybchenko <arybchenko at solarflare.com>
> >> ---
> > [...]
> >>   drivers/net/enic/enic.h                   |  2 +-
> >>   drivers/net/enic/enic_ethdev.c            | 22 +++++++---
> >>   drivers/net/enic/enic_main.c              |  4 +-
> > [...]
> >>   static void
> >> diff --git a/drivers/net/enic/enic.h b/drivers/net/enic/enic.h
> >> index 5a92508f00..72b1e7956b 100644
> >> --- a/drivers/net/enic/enic.h
> >> +++ b/drivers/net/enic/enic.h
> >> @@ -305,7 +305,7 @@ int enic_get_link_status(struct enic *enic);
> >>   int enic_dev_stats_get(struct enic *enic,
> >>   		       struct rte_eth_stats *r_stats);
> >>   void enic_dev_stats_clear(struct enic *enic);
> >> -void enic_add_packet_filter(struct enic *enic);
> >> +int enic_add_packet_filter(struct enic *enic);
> >>   int enic_set_mac_address(struct enic *enic, uint8_t *mac_addr);
> >>   int enic_del_mac_address(struct enic *enic, int mac_index);
> >>   unsigned int enic_cleanup_wq(struct enic *enic, struct vnic_wq *wq);
> >> diff --git a/drivers/net/enic/enic_ethdev.c
> b/drivers/net/enic/enic_ethdev.c
> >> index 90fdeda901..5d48930a9d 100644
> >> --- a/drivers/net/enic/enic_ethdev.c
> >> +++ b/drivers/net/enic/enic_ethdev.c
> >> @@ -603,29 +603,39 @@ static const uint32_t
> >> *enicpmd_dev_supported_ptypes_get(struct rte_eth_dev *dev)
> >>   	return NULL;
> >>   }
> >>
> >> -static void enicpmd_dev_promiscuous_enable(struct rte_eth_dev
> >> *eth_dev)
> >> +static int enicpmd_dev_promiscuous_enable(struct rte_eth_dev
> *eth_dev)
> >>   {
> >>   	struct enic *enic = pmd_priv(eth_dev);
> >> +	int ret;
> >>
> >>   	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
> >> -		return;
> >> +		return -ENOTSUP;
> > Should return -E_RTE_SECONDARY to be consistent with other handlers
> > that check primary/secondary.
> 
> I'll fix in the next version, but please, note that -ENOTSUP has
> special handling in ethdev patch on config restore. It looks like
> different error code should not be a problem here, but please, check.
> 

Andrew,

I see v3 returns -E_RTE_SECONDARY. Thank you. For enic, it does not
affect the new error checking in config_restore. In a secondary
process, enic's dev_start handler fails first, so rte_eth_dev_start
never gets to call config_restore.

-Hyong

> >>   	ENICPMD_FUNC_TRACE();
> >>
> >>   	enic->promisc = 1;
> >> -	enic_add_packet_filter(enic);
> >> +	ret = enic_add_packet_filter(enic);
> >> +	if (ret != 0)
> >> +		enic->promisc = 0;
> >> +
> >> +	return ret;
> >>   }
> >>
> >> -static void enicpmd_dev_promiscuous_disable(struct rte_eth_dev
> >> *eth_dev)
> >> +static int enicpmd_dev_promiscuous_disable(struct rte_eth_dev
> *eth_dev)
> >>   {
> >>   	struct enic *enic = pmd_priv(eth_dev);
> >> +	int ret;
> >>
> >>   	if (rte_eal_process_type() != RTE_PROC_PRIMARY)
> >> -		return;
> >> +		return -ENOTSUP;
> > Should return -E_RTE_SECONDARY here too.
> >
> >>   	ENICPMD_FUNC_TRACE();
> >>   	enic->promisc = 0;
> >> -	enic_add_packet_filter(enic);
> >> +	ret = enic_add_packet_filter(enic);
> >> +	if (ret != 0)
> >> +		enic->promisc = 1;
> >> +
> >> +	return ret;
> >>   }
> >>
> >>   static void enicpmd_dev_allmulticast_enable(struct rte_eth_dev
> *eth_dev)
> >> diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c
> >> index 40af3781b3..f4e76a057a 100644
> >> --- a/drivers/net/enic/enic_main.c
> >> +++ b/drivers/net/enic/enic_main.c
> >> @@ -1364,10 +1364,10 @@ int enic_set_vlan_strip(struct enic *enic)
> >>   			       enic->rss_enable);
> >>   }
> >>
> >> -void enic_add_packet_filter(struct enic *enic)
> >> +int enic_add_packet_filter(struct enic *enic)
> >>   {
> >>   	/* Args -> directed, multicast, broadcast, promisc, allmulti */
> >> -	vnic_dev_packet_filter(enic->vdev, 1, 1, 1,
> >> +	return vnic_dev_packet_filter(enic->vdev, 1, 1, 1,
> >>   		enic->promisc, enic->allmulti);
> >>   }
> >>
> > A couple minor comments above. Other than those, patch works fine for
> enic.
> > Feel free to add my ack on v2..
> >
> > Acked-by: Hyong Youb Kim <hyonkim at cisco.com>
> >
> 
> Thanks for review,
> Andrew.


More information about the dev mailing list