Re: [PATCH v3 2/2] IB/core: Add generic function to extract IB speed from netdev

[Yuval Shaia <yuval.shaia-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>]

On Wed, Jun 14, 2017 at 03:04:43PM +0000, Christian Benvenuti (benve) wrote:
> > -----Original Message-----
> > From: Yuval Shaia [mailto:yuval.shaia-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org]
> > Sent: Wednesday, June 14, 2017 6:03 AM
> > To: Christian Benvenuti (benve)
> > Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org; sean.hefty-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org; hal.rosenstock-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org;
> > selvin.xavier-dY08KVG/lbpWk0Htik3J/w@public.gmane.org; devesh.sharma-dY08KVG/lbpWk0Htik3J/w@public.gmane.org;
> > somnath.kotur-dY08KVG/lbpWk0Htik3J/w@public.gmane.org; sriharsha.basavapatna-dY08KVG/lbpWk0Htik3J/w@public.gmane.org; Dave Goodell
> > (dgoodell); monis-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org; leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org; ira.weiny-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org;
> > dasaratharaman.chandramouli-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org; sagi-TmH2Wj2nsNJBDLzU/O5InQ@public.gmane.org;
> > bart.vanassche-XdAiOPVOjttBDgjK7y7TUQ@public.gmane.org; yishaih-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org; linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > Subject: Re: [PATCH v3 2/2] IB/core: Add generic function to extract IB speed
> > from netdev
> > 
> > On Tue, Jun 13, 2017 at 08:07:48AM +0000, Christian Benvenuti (benve) wrote:
> > > >
> > > > +int ib_get_eth_speed(struct ib_device *dev, u8 port_num, u8 *speed, u8
> > > > +*width) {
> > > > +	int rc;
> > > > +	u32 netdev_speed;
> > > > +	struct net_device *netdev;
> > > > +	struct ethtool_link_ksettings lksettings;
> > > > +
> > > > +	if (rdma_port_get_link_layer(dev, port_num) != IB_LINK_LAYER_ETHERNET)
> > > > +		return -EINVAL;
> > > > +
> > > > +	if (!dev->get_netdev)
> > > > +		return -EINVAL;
> > >
> > > -EOPNOTSUPP ?
> > 
> > Looks better.
> > Assuming the same should be for the above one and the one below, right?
> 
> I am not sure about the other two:
> - rdma_port_get_link_layer() does not entirely depend on ib_device->get_link_layer()
> - ib_device->get_netdev() can fail for different reasons I guess

I see,

So back to your proposal, the fact that dev->get_netdev is not set means
that the IB device is not relayed on Ethernet one (RoCE folks, correct me
if i'm wrong) so with the given input - the operation is not supported. Is
that means that -EINVAL is better than -EOPNOTSUPP?

As far as the name suggest, EINVAL is where input values are wrong while
EOPNOTSUPP is where input values are fine but the operation is not
supported.

Yuval

> 
> /Chris
> 
> > >
> > > > +	netdev = dev->get_netdev(dev, port_num);
> > > > +	if (!netdev)
> > > > +		return -EINVAL;
> > > > +
> > > > +	rtnl_lock();
> > > > +	rc = __ethtool_get_link_ksettings(netdev, &lksettings);
> > > > +	rtnl_unlock();
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html