On Tue, Jun 24, 2008 at 11:56 PM, Roland Dreier <[EMAIL PROTECTED]> wrote:
> To be honest this looks like a step backwards...
>
> > +static ssize_t show_use_lro(struct device *d, struct device_attribute
> *attr,
> > + char *buf)
> > +{
> > + struct ipoib_dev_priv *priv = netdev_priv(to_net_dev(d));
> > +
> > + if (priv->lro.use_lro)
> > + return sprintf(buf, "yes\n");
> > + else
> > + return sprintf(buf, "no\n");
> > +}
>
> This looks like a crazy interface ("yes"/"no" vs 1/0??) and in any case
> I think using the existing ethtool support to control LRO would be much
> simpler and better for users.
>
> - R.
>
>
I can change "yes" to "1" and "no" to "0".
I don't see that there is LRO support in the userspace ethtool.
Regards,
VLadimir
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
