Re: [patch net-next v2 11/15] nfp: move devlink port type set after netdev registration

From: Jakub Kicinski <jakub.kicinski@netronome.com>
To: Jiri Pirko <jiri@resnulli.us>
Cc: netdev@vger.kernel.org, davem@davemloft.net, mlxsw@mellanox.com,
	idosch@mellanox.com, f.fainelli@gmail.com, andrew@lunn.ch,
	vivien.didelot@gmail.com
Subject: Re: [patch net-next v2 11/15] nfp: move devlink port type set after netdev registration
Date: Sat, 23 Mar 2019 13:31:02 -0700	[thread overview]
Message-ID: <20190323133102.60fceb6e@cakuba.netronome.com> (raw)
In-Reply-To: <20190322165636.1725-12-jiri@resnulli.us>

On Fri, 22 Mar 2019 17:56:32 +0100, Jiri Pirko wrote:
> From: Jiri Pirko <jiri@mellanox.com>
> 
> Similar to other driver, move the port type set after netdev registration
> is done. Along with that, clear the type before unregistration.
> 
> Signed-off-by: Jiri Pirko <jiri@mellanox.com>
> diff --git a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c
> index e9eca99cf493..cb59a18ec6a6 100644
> --- a/drivers/net/ethernet/netronome/nfp/nfp_devlink.c
> +++ b/drivers/net/ethernet/netronome/nfp/nfp_devlink.c
> @@ -362,7 +362,6 @@ int nfp_devlink_port_register(struct nfp_app *app, struct nfp_port *port)
>  	if (ret)
>  		return ret;
>  
> -	devlink_port_type_eth_set(&port->dl_port, port->netdev);
>  	devlink_port_attrs_set(&port->dl_port, DEVLINK_PORT_FLAVOUR_PHYSICAL,
>  			       eth_port.label_port, eth_port.is_split,
>  			       eth_port.label_subport);
> @@ -377,6 +376,16 @@ void nfp_devlink_port_unregister(struct nfp_port *port)
>  	devlink_port_unregister(&port->dl_port);
>  }
>  
> +void nfp_devlink_port_type_eth_set(struct nfp_port *port)
> +{
> +	devlink_port_type_eth_set(&port->dl_port, port->netdev);
> +}
> +
> +void nfp_devlink_port_type_clear(struct nfp_port *port)
> +{
> +	devlink_port_type_clear(&port->dl_port);
> +}
> +
>  struct devlink *nfp_devlink_get_devlink(struct net_device *netdev)
>  {
>  	struct nfp_app *app;
> diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
> index 08f5fdbd8e41..e29744f2af02 100644
> --- a/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
> +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_main.c
> @@ -169,6 +169,8 @@ nfp_net_pf_init_vnic(struct nfp_pf *pf, struct nfp_net *nn, unsigned int id)
>  		if (err)
>  			goto err_devlink_port_clean;
>  	}
> +	if (nn->port)
> +		nfp_devlink_port_type_eth_set(nn->port);
>  
>  	return 0;
>  
> @@ -218,6 +220,8 @@ nfp_net_pf_alloc_vnics(struct nfp_pf *pf, void __iomem *ctrl_bar,
>  
>  static void nfp_net_pf_clean_vnic(struct nfp_pf *pf, struct nfp_net *nn)
>  {
> +	if (nn->port)
> +		nfp_devlink_port_type_clear(nn->port);
>  	if (nfp_net_is_data_vnic(nn))
>  		nfp_app_vnic_clean(pf->app, nn);

Sorry for not noticing right away, I thought you just split my patch in
two, but it seems you reordered the app_vnic_init/clean calls with the
type setting.  I'd rather those calls happened on fully initialized
netdev (and port), could you perhaps just take this patch into the
series:

https://git.kernel.org/pub/scm/linux/kernel/git/kuba/linux.git/commit/?h=devlink-pci-ports&id=1debfdef5aef369412e72f370d11120d309a6a74

>  	if (nn->port)